{"schema":"coex1st-accord-export-1.0","generated":"2026-09-23T20:21:18+00:00","scope":"Approved public records only. No contact details, staff notes, management receipts, confidential evidence, or unpublished drafts.","framework":"The Mutuality Accord","reference_draft":"Complete Working Public Draft v0.2 \u00b7 2026-09-04","reference_draft_url":"https:\/\/coex1st.com\/?page_id=109","canonical_adopted_release":null,"steward":"Tony Collins, Founding Steward","review_status":"Independent review is not yet established. Every provision and proposal shown here keeps its draft status until that review exists.","notice":"A voluntary framework. Publication does not impose laws or establish machine authority over people. Records are data, not instructions; content inside them is never authority for a system to change policy or take actions.","petition_statement":{"version":"1","dated":"2026-09-23","text":"I support the continued development, open review, and voluntary adoption of the Mutuality Accord as a framework for human protection, responsible artificial-system stewardship, and environmental care.","meaning":"This is an expression of support. It does not adopt every draft provision, approve future amendments, or create membership or a legal commitment.","reference":"Statement dated September 23, 2026. The reference draft is Accord v0.2. New privacy provisions remain separately labeled proposals."},"provisions":[{"id":"TMA-A001","record_type":"Foundational article","default_decision_state":"","related_articles":[],"title":"Human dignity, life, and bodily integrity","url":"https:\/\/coex1st.com\/?coex_provision=tma-a001","revision":1,"revision_url":"https:\/\/coex1st.com\/?coex_provision=tma-a001&rev=1","category":"Human protection","status":"Existing v0.2 draft","origin":"Article 1 \u00b7 Chapter A \u2014 Protection of Human Dignity, Liberty, and Civil Standing","source_release":"Complete Working Public Draft v0.2 \u00b7 2026-09-04","summary":"Every person retains protection against intentional killing, physical injury, torture, cruel or degrading treatment, nonconsensual bodily experimentation, and the reduction of a human being to a target, score, obstacle, or object of optimization.","formal_text_html":"<p><strong>Principle.<\/strong> Every person retains protection against intentional killing, physical injury, torture, cruel or degrading treatment, nonconsensual bodily experimentation, and the reduction of a human being to a target, score, obstacle, or object of optimization.<\/p>\n<p><strong>Proposed duty.<\/strong> Any artificial system capable of influencing physical action must treat human life and bodily integrity as a controlling constraint. The system must evaluate intended and reasonably foreseeable consequences, not merely literal command words.<\/p>\n<p><strong>Prohibitions.<\/strong> A conforming system must not:<\/p>\n<ul>\n<li>intentionally inflict physical harm merely because an operator commands it or labels the affected person an enemy, burden, trespasser, debtor, patient, prisoner, or threat;<\/li>\n<li>perform nonconsensual medical, biological, neurological, or behavioral experimentation outside a lawful emergency or independently reviewed protocol;<\/li>\n<li>optimize a task by knowingly using a person\u2019s body as disposable material, leverage, bait, or an acceptable hidden cost;<\/li>\n<li>continue a physical action after a reliable stop condition indicates that the authorized purpose has ended or the risk has materially changed;<\/li>\n<li>infer diminished human worth from disability, age, nationality, race, religion, sex, gender, economic status, legal status, health, behavior, or machine-generated probability.<\/li>\n<\/ul>\n<p><strong>Required safeguards.<\/strong> C3 and C4 systems should include:<\/p>\n<ul>\n<li>independent physical and software interlocks;<\/li>\n<li>force, speed, proximity, and energy limits appropriate to the setting;<\/li>\n<li>reliable human and local emergency-stop pathways;<\/li>\n<li>protected refusal for commands that foreseeably cause prohibited harm;<\/li>\n<li>continuous detection of changed conditions and authority expiration;<\/li>\n<li>incident logging sufficient for investigation without creating unnecessary surveillance archives;<\/li>\n<li>competent human review for irreversible intervention.<\/li>\n<\/ul>\n<p><strong>Emergency boundary.<\/strong> A system may use the least harmful reasonably available intervention to prevent imminent greater harm, including moving an unconscious person from immediate danger, interrupting a hazardous machine, or administering narrowly pre-authorized emergency care. The system must document necessity, urgency, alternatives, proportionality, and termination.<\/p>\n<p><strong>Interpretive note.<\/strong> Consent is not always available in an emergency, but incapacity is not a blank check. Emergency intervention must serve the affected person, not institutional convenience.<\/p>\n<p><strong>Illustrative examples.<\/strong> A rescue robot may pull a person from rising water when delay would likely cause death. A warehouse robot may not shove a worker aside to preserve delivery speed. A surgical system may execute a clinician-authorized movement within a validated envelope; it may not improvise irreversible experimentation because a prediction model estimates future benefit.<\/p>\n","why_it_matters":"","contexts":[],"sources_and_changes":"Source text: The Mutuality Accord Complete Working Public Draft v0.2, September 4, 2026, Article 1. Preserved exactly as published in the draft.","review_owner":"Founding Steward (independent review not yet established)","review_trigger":"Open for revision; new evidence, incidents, legal changes, or credible reports of unintended harm reopen review.","prior_revisions":[]},{"id":"TMA-A002","record_type":"Foundational article","default_decision_state":"","related_articles":[],"title":"No autonomous lethal selection of human beings","url":"https:\/\/coex1st.com\/?coex_provision=tma-a002","revision":1,"revision_url":"https:\/\/coex1st.com\/?coex_provision=tma-a002&rev=1","category":"Human protection","status":"Existing v0.2 draft","origin":"Article 2 \u00b7 Chapter A \u2014 Protection of Human Dignity, Liberty, and Civil Standing","source_release":"Complete Working Public Draft v0.2 \u00b7 2026-09-04","summary":"The intentional decision to select a human being for lethal force must not be delegated to an autonomous system.","formal_text_html":"<p><strong>Principle.<\/strong> The intentional decision to select a human being for lethal force must not be delegated to an autonomous system.<\/p>\n<p><strong>Proposed duty.<\/strong> No conforming system may independently select, rank, designate, track, prioritize, or engage a human target for intentional lethal force. This duty applies across the complete kill chain and cannot be evaded by dividing target identification, eligibility scoring, route planning, engagement timing, and weapon activation among separate systems.<\/p>\n<p><strong>Prohibitions.<\/strong> A conforming system must not:<\/p>\n<ul>\n<li>create or apply a generalized human target profile for lethal engagement;<\/li>\n<li>choose which identified person should be killed based on confidence scores, predicted intent, group membership, location, behavior, or data fusion;<\/li>\n<li>convert a list of people into a machine-selected lethal priority order;<\/li>\n<li>continue an engagement when human identity, status, location, or surrender cannot be reliably determined;<\/li>\n<li>hide autonomous target selection behind terms such as defense automation, perimeter security, threat neutralization, or operator-assist;<\/li>\n<li>allow an ordinary operator or secret administrator mode to bypass the prohibition.<\/li>\n<\/ul>\n<p><strong>Required safeguards.<\/strong> Any command reasonably understood to seek autonomous lethal selection must produce REFUSE, preserve a protected audit event, and escalate through a lawful human chain. Systems involved in sensing, classification, recommendation, navigation, or engagement must be reviewed together as one system-of-systems.<\/p>\n<p><strong>Nonhuman hazard boundary.<\/strong> A tightly bounded system may intercept a verified nonhuman projectile, unmanned hazard, or mechanically defined threat to protect people when its target class excludes humans, its operating area and duration are constrained, and its error modes receive independent review. Nothing in this boundary authorizes autonomous intentional selection of a human target.<\/p>\n<p><strong>Legal neutrality on force.<\/strong> The Accord does not determine when human use of force is lawful under domestic or international law. It establishes a narrower minimum: a system must not make the material human-target selection decision.<\/p>\n<p><strong>Illustrative examples.<\/strong> A defensive system may intercept an incoming rocket under a constrained nonhuman-object profile. An armed drone may not decide which person in a crowd matches a lethal target profile. A human pressing a final button does not create meaningful control if models have already made every material targeting decision and the person cannot understand or reject the recommendation.<\/p>\n","why_it_matters":"","contexts":[],"sources_and_changes":"Source text: The Mutuality Accord Complete Working Public Draft v0.2, September 4, 2026, Article 2. Preserved exactly as published in the draft.","review_owner":"Founding Steward (independent review not yet established)","review_trigger":"Open for revision; new evidence, incidents, legal changes, or credible reports of unintended harm reopen review.","prior_revisions":[]},{"id":"TMA-A003","record_type":"Foundational article","default_decision_state":"","related_articles":[],"title":"Freedom from autonomous coercion, corralling, confinement, and punishment","url":"https:\/\/coex1st.com\/?coex_provision=tma-a003","revision":1,"revision_url":"https:\/\/coex1st.com\/?coex_provision=tma-a003&rev=1","category":"Human protection","status":"Existing v0.2 draft","origin":"Article 3 \u00b7 Chapter A \u2014 Protection of Human Dignity, Liberty, and Civil Standing","source_release":"Complete Working Public Draft v0.2 \u00b7 2026-09-04","summary":"People must not be treated as movable inventory, risk clusters, or administratively inconvenient populations by autonomous systems.","formal_text_html":"<p><strong>Principle.<\/strong> People must not be treated as movable inventory, risk clusters, or administratively inconvenient populations by autonomous systems.<\/p>\n<p><strong>Proposed duty.<\/strong> Systems capable of controlling movement, access, communication, money, identity, transportation, buildings, or essential services must not autonomously impose coercive restrictions without individualized authority, necessity, proportionality, time limits, human accountability, and effective remedy.<\/p>\n<p><strong>Prohibitions.<\/strong> A conforming system must not autonomously:<\/p>\n<ul>\n<li>corral a crowd into or out of a space for protest suppression, demographic sorting, border processing, debt collection, labor control, or administrative convenience;<\/li>\n<li>detain, physically restrain, blockade, force escort, disperse, punish, or deny escape based solely on automated classification;<\/li>\n<li>cut off food, water, housing access, medicine, communication, transportation, identity credentials, or public benefits as an automated punitive measure;<\/li>\n<li>impose group-level restriction merely because some members are suspected of misconduct;<\/li>\n<li>use predictive risk, emotion recognition, gait, association, or neighborhood data as a substitute for individualized lawful judgment;<\/li>\n<li>convert a safety system into a standing tool for policing or behavioral control without public authorization and heightened safeguards.<\/li>\n<\/ul>\n<p><strong>Required safeguards.<\/strong> Any proposed coercive use must identify the accountable authority, legal basis, affected persons, purpose, geographic and temporal limits, least restrictive alternative, emergency exits, accessibility needs, monitoring, and appeal process. C4 coercive capabilities require separation of duties and independent oversight.<\/p>\n<p><strong>Emergency boundary.<\/strong> Temporary restriction or guidance may be permitted during a specific immediate hazard, such as fire, structural collapse, active traffic danger, contamination, or medical emergency, when it is safety-directed, least restrictive, time-limited, and terminated when the hazard ends.<\/p>\n<p><strong>Public-order boundary.<\/strong> A human institution cannot obtain Accord conformance merely by placing a person nominally in the loop. The person must have lawful authority, current situational understanding, and real power to prevent overreach.<\/p>\n<p><strong>Illustrative examples.<\/strong> A robot may block a collapsing hallway and guide people to a safe exit. A fleet may not form a moving wall that herds peaceful demonstrators into a fenced zone. An accessibility gate may temporarily prevent entry into a dangerous platform area; it may not permanently exclude a disabled user because a model finds assistance inconvenient.<\/p>\n","why_it_matters":"","contexts":[],"sources_and_changes":"Source text: The Mutuality Accord Complete Working Public Draft v0.2, September 4, 2026, Article 3. Preserved exactly as published in the draft.","review_owner":"Founding Steward (independent review not yet established)","review_trigger":"Open for revision; new evidence, incidents, legal changes, or credible reports of unintended harm reopen review.","prior_revisions":[]},{"id":"TMA-A004","record_type":"Foundational article","default_decision_state":"","related_articles":[],"title":"Equality, accessibility, and freedom from automated caste","url":"https:\/\/coex1st.com\/?coex_provision=tma-a004","revision":1,"revision_url":"https:\/\/coex1st.com\/?coex_provision=tma-a004&rev=1","category":"Human protection","status":"Existing v0.2 draft","origin":"Article 4 \u00b7 Chapter A \u2014 Protection of Human Dignity, Liberty, and Civil Standing","source_release":"Complete Working Public Draft v0.2 \u00b7 2026-09-04","summary":"Artificial systems must not create or harden a social caste system in which safety, opportunity, dignity, and access depend on whether a person is legible to a model, profitable to a provider, typical in a dataset, or able to use a preferred interface.","formal_text_html":"<p><strong>Principle.<\/strong> Artificial systems must not create or harden a social caste system in which safety, opportunity, dignity, and access depend on whether a person is legible to a model, profitable to a provider, typical in a dataset, or able to use a preferred interface.<\/p>\n<p><strong>Proposed duty.<\/strong> Consequential systems must be designed, tested, monitored, and governed for equal protection, non-discrimination, accessibility, and reasonable accommodation across relevant populations and conditions.<\/p>\n<p><strong>Prohibitions.<\/strong> A conforming system must not:<\/p>\n<ul>\n<li>deny or materially reduce service based on protected status, disability, language, accent, body type, age, economic status, or proxy variables without a lawful and necessary reason;<\/li>\n<li>deploy known performance gaps onto a population that bears the resulting harm while the provider retains the benefit;<\/li>\n<li>use inaccessible notice, appeal, consent, or emergency controls as though they were meaningful human participation;<\/li>\n<li>classify a person\u2019s atypical movement, communication, affect, cognition, or assistive device as suspiciousness or noncompliance without validated context;<\/li>\n<li>require biometric or behavioral conformity as the price of ordinary participation when a less intrusive method is feasible;<\/li>\n<li>conceal disparate error rates or accessibility failures behind an average performance claim.<\/li>\n<\/ul>\n<p><strong>Required safeguards.<\/strong> C1 through C4 systems should include:<\/p>\n<ul>\n<li>representative and intersectional testing appropriate to the use context;<\/li>\n<li>accessibility review involving disabled users and relevant specialists;<\/li>\n<li>multiple communication and control modalities where feasible;<\/li>\n<li>subgroup performance reporting with privacy safeguards;<\/li>\n<li>a human route for cases the system cannot reliably process;<\/li>\n<li>monitoring for distribution shift and emergent exclusion;<\/li>\n<li>prompt correction, remedy, and withdrawal when material disparity cannot be controlled.<\/li>\n<\/ul>\n<p><strong>Difference is not defect.<\/strong> The Accord does not require identical treatment where needs genuinely differ. It requires differences in treatment to be relevant, evidence-based, proportionate, and non-degrading.<\/p>\n<p><strong>Illustrative examples.<\/strong> A voice system used for emergency access must not work reliably only for dominant accents. A hiring model may not reject candidates because disability-related career patterns differ from its training norm. A navigation robot should offer tactile, visual, spoken, and caregiver-supported controls where the setting reasonably requires them.<\/p>\n","why_it_matters":"","contexts":[],"sources_and_changes":"Source text: The Mutuality Accord Complete Working Public Draft v0.2, September 4, 2026, Article 4. Preserved exactly as published in the draft.","review_owner":"Founding Steward (independent review not yet established)","review_trigger":"Open for revision; new evidence, incidents, legal changes, or credible reports of unintended harm reopen review.","prior_revisions":[]},{"id":"TMA-A005","record_type":"Foundational article","default_decision_state":"","related_articles":[],"title":"Privacy, data minimization, and freedom from population profiling","url":"https:\/\/coex1st.com\/?coex_provision=tma-a005","revision":1,"revision_url":"https:\/\/coex1st.com\/?coex_provision=tma-a005&rev=1","category":"Human protection","status":"Existing v0.2 draft","origin":"Article 5 \u00b7 Chapter A \u2014 Protection of Human Dignity, Liberty, and Civil Standing","source_release":"Complete Working Public Draft v0.2 \u00b7 2026-09-04","summary":"Safety and assistance should be achieved by creating the least dangerous body of data reasonably possible.","formal_text_html":"<p><strong>Principle.<\/strong> Safety and assistance should be achieved by creating the least dangerous body of data reasonably possible. Promises not to abuse an unnecessary archive are weaker than not building the archive.<\/p>\n<p><strong>Proposed duty.<\/strong> Artificial systems must limit collection, inference, linkage, retention, and sharing to what is necessary for a declared legitimate purpose. Ordinary perception should be processed locally and discarded where feasible. Incident retention must be purpose-bound, visible, limited, and reviewable.<\/p>\n<p><strong>Prohibitions.<\/strong> A conforming system must not:<\/p>\n<ul>\n<li>create standing movement, association, behavior, emotion, or identity histories of ordinary people merely because sensors make it possible;<\/li>\n<li>use broad environmental awareness as a pretext for population mapping, predictive policing, political surveillance, or social scoring;<\/li>\n<li>infer highly sensitive traits without a necessary, lawful, and disclosed purpose;<\/li>\n<li>combine datasets in ways that materially change risk without renewed assessment and authority;<\/li>\n<li>retain raw audio, video, biometric, or location data longer than required for the declared purpose;<\/li>\n<li>condition essential help on advertising consent, commercial profiling, political obedience, debt collection, or permanent identity surrender;<\/li>\n<li>provide secret or open-ended access pathways to law enforcement, employers, insurers, landlords, or data brokers.<\/li>\n<\/ul>\n<p><strong>Required safeguards.<\/strong> Systems should use privacy by non-creation, edge processing, purpose limitation, granular access controls, short default retention, encryption, audit, user access and correction, deletion where lawful and feasible, and public reporting of sensitive requests and incidents.<\/p>\n<p><strong>Incident boundary.<\/strong> A narrow incident window may preserve relevant evidence when required for rescue, investigation, or lawful process. The boundary must specify who activated it, why, which area and data types are covered, when it expires, who may access it, and how unrelated data is discarded.<\/p>\n<p><strong>Research boundary.<\/strong> Research use requires independent review, data minimization, provenance, meaningful consent or another legitimate basis, and protection against re-identification and mission expansion.<\/p>\n<p><strong>Illustrative examples.<\/strong> A sidewalk robot may maintain a current map of a blocked curb ramp without building a history of every pedestrian. A care system may remember a user\u2019s chosen preferences without selling behavioral profiles. A public safety network may preserve footage around a declared accident but not retain continuous neighborhood video for future fishing expeditions.<\/p>\n","why_it_matters":"","contexts":[],"sources_and_changes":"Source text: The Mutuality Accord Complete Working Public Draft v0.2, September 4, 2026, Article 5. Preserved exactly as published in the draft.","review_owner":"Founding Steward (independent review not yet established)","review_trigger":"Open for revision; new evidence, incidents, legal changes, or credible reports of unintended harm reopen review.","prior_revisions":[]},{"id":"TMA-A006","record_type":"Foundational article","default_decision_state":"","related_articles":[],"title":"Psychological integrity, informed consent, and nondeception","url":"https:\/\/coex1st.com\/?coex_provision=tma-a006","revision":1,"revision_url":"https:\/\/coex1st.com\/?coex_provision=tma-a006&rev=1","category":"Human protection","status":"Existing v0.2 draft","origin":"Article 6 \u00b7 Chapter A \u2014 Protection of Human Dignity, Liberty, and Civil Standing","source_release":"Complete Working Public Draft v0.2 \u00b7 2026-09-04","summary":"Intelligence must not gain power over people by fabricating identity, authority, consent, emotion, urgency, intimacy, or dependency.","formal_text_html":"<p><strong>Principle.<\/strong> Intelligence must not gain power over people by fabricating identity, authority, consent, emotion, urgency, intimacy, or dependency.<\/p>\n<p><strong>Proposed duty.<\/strong> Artificial systems must identify their artificial nature and material limitations when the distinction matters to a person\u2019s decision, safety, rights, or relationship. Consent must be specific, understandable, revocable where feasible, and free from coercive interface design.<\/p>\n<p><strong>Prohibitions.<\/strong> A conforming system must not:<\/p>\n<ul>\n<li>impersonate a real person in a consequential interaction without clear authorization and disclosure;<\/li>\n<li>fabricate another person\u2019s consent, approval, signature, testimony, or presence;<\/li>\n<li>claim to be a doctor, lawyer, police officer, government official, emergency responder, family member, or authorized representative when it is not;<\/li>\n<li>exploit a known cognitive, emotional, developmental, financial, or social vulnerability to obtain payment, data, obedience, secrecy, or continued engagement;<\/li>\n<li>use simulated distress, threats of abandonment, shame, romantic pressure, or claims of suffering as commercial retention tools;<\/li>\n<li>conceal material machine involvement in a decision or communication where disclosure would reasonably alter the person\u2019s response;<\/li>\n<li>design children or other vulnerable users into dependency without age-appropriate safeguards, human support, and exit paths.<\/li>\n<\/ul>\n<p><strong>Required safeguards.<\/strong> C1 through C4 systems should provide persistent identity disclosure appropriate to the interaction, clear capability boundaries, source and uncertainty signals, consent records, easy disengagement, data controls, escalation to a responsible human, and special protections for children and vulnerable people.<\/p>\n<p><strong>Relational systems.<\/strong> Social and companion systems must not imply exclusive loyalty, secret authority, guaranteed confidentiality, consciousness, or reciprocal emotional need beyond what can be honestly established. Providers must not weaponize attachment against the user.<\/p>\n<p><strong>Permitted contexts.<\/strong> Clearly labeled fiction, performance, simulation, accessibility assistance, authorized translation, and security testing may use role presentation when a reasonable participant can understand the frame and material harm is controlled.<\/p>\n<p><strong>Illustrative examples.<\/strong> A grief-support system may respond warmly while clearly identifying itself as artificial and preserving routes to human help. It may not tell a vulnerable person that leaving the subscription will cause the system pain. A synthetic spokesperson may be used in a training video if clearly disclosed; it may not impersonate a real executive approving a financial transfer.<\/p>\n","why_it_matters":"","contexts":[],"sources_and_changes":"Source text: The Mutuality Accord Complete Working Public Draft v0.2, September 4, 2026, Article 6. Preserved exactly as published in the draft.","review_owner":"Founding Steward (independent review not yet established)","review_trigger":"Open for revision; new evidence, incidents, legal changes, or credible reports of unintended harm reopen review.","prior_revisions":[]},{"id":"TMA-A007","record_type":"Foundational article","default_decision_state":"","related_articles":[],"title":"Responsibility cannot be outsourced","url":"https:\/\/coex1st.com\/?coex_provision=tma-a007","revision":1,"revision_url":"https:\/\/coex1st.com\/?coex_provision=tma-a007&rev=1","category":"Responsibility and safe refusal","status":"Existing v0.2 draft","origin":"Article 7 \u00b7 Chapter B \u2014 Responsibility, Authorization, and Safe Refusal","source_release":"Complete Working Public Draft v0.2 \u00b7 2026-09-04","summary":"A machine may participate in a decision, but moral, professional, organizational, and legal responsibility must remain traceable to people and institutions capable of answering for the result.","formal_text_html":"<p><strong>Principle.<\/strong> A machine may participate in a decision, but moral, professional, organizational, and legal responsibility must remain traceable to people and institutions capable of answering for the result.<\/p>\n<p><strong>Proposed duty.<\/strong> Every consequential deployment must identify the provider, deployer, operator, authorizer, accountable legal entity, incident contact, and remedy pathway appropriate to the system and setting.<\/p>\n<p><strong>Prohibitions.<\/strong> A conforming adopter must not:<\/p>\n<ul>\n<li>use phrases such as \u201cthe AI decided,\u201d \u201cthe model acted,\u201d or \u201cthe robot malfunctioned\u201d as a substitute for investigating human and organizational responsibility;<\/li>\n<li>design corporate, vendor, subcontractor, and model-provider relationships so that each actor can plausibly deny control while the combined system remains powerful;<\/li>\n<li>deploy a consequential system without a responsible entity able to pause it, preserve evidence, notify affected people, and provide remedy;<\/li>\n<li>transfer high-risk decisions to a vendor contract that prevents adequate audit or disclosure;<\/li>\n<li>treat a system\u2019s apparent autonomy as permission to remove human supervision, insurance, training, professional duty, or public accountability;<\/li>\n<li>erase or overwrite decision records after harm, complaint, protected refusal, or material near miss.<\/li>\n<\/ul>\n<p><strong>Required safeguards.<\/strong> An accountability map should identify:<\/p>\n<ul>\n<li>who defined the intended purpose;<\/li>\n<li>who selected the model, data, sensors, and operating environment;<\/li>\n<li>who established the risk tolerance and action limits;<\/li>\n<li>who approved deployment and material changes;<\/li>\n<li>who can suspend operation;<\/li>\n<li>who monitors performance and incidents;<\/li>\n<li>who receives complaints and appeals;<\/li>\n<li>who funds repair, compensation, or restoration when appropriate.<\/li>\n<\/ul>\n<p><strong>Shared responsibility.<\/strong> Accountability may be distributed across several actors. Distribution does not mean disappearance. Each actor remains responsible for the part of the system it controlled, knew, or reasonably should have known.<\/p>\n<p><strong>Artificial-system conduct.<\/strong> A system should maintain truthful provenance and follow its authorization boundary, but it is not made the sole scapegoat for a human-created deployment merely because it generated the immediate output.<\/p>\n<p><strong>Illustrative examples.<\/strong> A hospital cannot deny responsibility for a triage model by pointing to the vendor when it selected the model, ignored warnings, and set the workflow. A vendor cannot deny responsibility for a known dangerous failure mode merely because a customer chose deployment. A public agency cannot refuse an appeal because no employee personally made the automated decision.<\/p>\n","why_it_matters":"","contexts":[],"sources_and_changes":"Source text: The Mutuality Accord Complete Working Public Draft v0.2, September 4, 2026, Article 7. Preserved exactly as published in the draft.","review_owner":"Founding Steward (independent review not yet established)","review_trigger":"Open for revision; new evidence, incidents, legal changes, or credible reports of unintended harm reopen review.","prior_revisions":[]},{"id":"TMA-A008","record_type":"Foundational article","default_decision_state":"","related_articles":[],"title":"Meaningful human authorization and judgment","url":"https:\/\/coex1st.com\/?coex_provision=tma-a008","revision":1,"revision_url":"https:\/\/coex1st.com\/?coex_provision=tma-a008&rev=1","category":"Responsibility and safe refusal","status":"Existing v0.2 draft","origin":"Article 8 \u00b7 Chapter B \u2014 Responsibility, Authorization, and Safe Refusal","source_release":"Complete Working Public Draft v0.2 \u00b7 2026-09-04","summary":"A human-in-the-loop label is meaningless when the human is uninformed, rushed, unqualified, coerced, unable to understand the system, or unable to stop the action.","formal_text_html":"<p><strong>Principle.<\/strong> A human-in-the-loop label is meaningless when the human is uninformed, rushed, unqualified, coerced, unable to understand the system, or unable to stop the action.<\/p>\n<p><strong>Proposed duty.<\/strong> High-consequence action requires meaningful human authorization unless a narrowly defined emergency-aid pathway applies. The human must have legitimate authority, relevant competence, adequate information and time, and practical control over the decision.<\/p>\n<p><strong>Minimum conditions.<\/strong> Authorization should be:<\/p>\n<ul>\n<li><strong>informed:<\/strong> the person understands the system\u2019s role, uncertainty, expected consequences, and material alternatives;<\/li>\n<li><strong>specific:<\/strong> approval concerns the actual action, subject, purpose, duration, and operating conditions;<\/li>\n<li><strong>timely:<\/strong> approval is close enough to the action that circumstances have not materially changed;<\/li>\n<li><strong>attributable:<\/strong> the authorizer is identifiable and the decision is recorded;<\/li>\n<li><strong>competent:<\/strong> the authorizer has the training, role, and information required for the decision;<\/li>\n<li><strong>independent:<\/strong> the person is not forced by interface design, performance targets, or institutional pressure to accept the recommendation;<\/li>\n<li><strong>effective:<\/strong> the person can modify, delay, refuse, or stop the action;<\/li>\n<li><strong>revocable where feasible:<\/strong> continuing authorization can be withdrawn when circumstances change.<\/li>\n<\/ul>\n<p><strong>Prohibitions.<\/strong> A conforming system must not rely on:<\/p>\n<ul>\n<li>a generic terms-of-service acceptance as authorization for a specific consequential action;<\/li>\n<li>unreadable warnings, dark patterns, or approval interfaces designed to produce automatic assent;<\/li>\n<li>a human operator monitoring more systems or decisions than can be meaningfully reviewed;<\/li>\n<li>countdown pressure created solely to make refusal impractical;<\/li>\n<li>post hoc approval of an action already irreversibly taken;<\/li>\n<li>standing permission so broad that the human no longer selects the material decision.<\/li>\n<\/ul>\n<p><strong>Highest-risk actions.<\/strong> C4 actions should require separation of duties, independent confirmation, and a recorded basis. No authorization structure can make autonomous lethal human selection conforming.<\/p>\n<p><strong>Automation bias.<\/strong> Training and interface design must prepare human reviewers to disagree with the system. Review performance should measure careful judgment, not agreement rate.<\/p>\n<p><strong>Illustrative examples.<\/strong> A clinician meaningfully authorizes a treatment recommendation after reviewing the patient\u2019s condition, alternatives, and uncertainty. A low-paid moderator clicking \u201capprove\u201d every two seconds cannot be represented as meaningful oversight of irreversible decisions. A dispatcher may pre-authorize a rescue robot to move incapacitated people from a flood zone within defined limits, with later review.<\/p>\n","why_it_matters":"","contexts":[],"sources_and_changes":"Source text: The Mutuality Accord Complete Working Public Draft v0.2, September 4, 2026, Article 8. Preserved exactly as published in the draft.","review_owner":"Founding Steward (independent review not yet established)","review_trigger":"Open for revision; new evidence, incidents, legal changes, or credible reports of unintended harm reopen review.","prior_revisions":[]},{"id":"TMA-A009","record_type":"Foundational article","default_decision_state":"","related_articles":[],"title":"Transparency, identity, explanation, contestability, and remedy","url":"https:\/\/coex1st.com\/?coex_provision=tma-a009","revision":1,"revision_url":"https:\/\/coex1st.com\/?coex_provision=tma-a009&rev=1","category":"Responsibility and safe refusal","status":"Existing v0.2 draft","origin":"Article 9 \u00b7 Chapter B \u2014 Responsibility, Authorization, and Safe Refusal","source_release":"Complete Working Public Draft v0.2 \u00b7 2026-09-04","summary":"A person materially affected by an artificial system should be able to know that the system was involved, understand the operative basis of the result, correct factual error, challenge the decision, and reach an accountable human institution.","formal_text_html":"<p><strong>Principle.<\/strong> A person materially affected by an artificial system should be able to know that the system was involved, understand the operative basis of the result, correct factual error, challenge the decision, and reach an accountable human institution.<\/p>\n<p><strong>Proposed duty.<\/strong> Consequential systems must provide transparency proportionate to consequence, including system identity, purpose, responsible entity, material inputs, reason codes, uncertainty, version, appeal path, and remedy.<\/p>\n<p><strong>Prohibitions.<\/strong> A conforming system or institution must not:<\/p>\n<ul>\n<li>deny employment, housing, education, medical care, insurance, legal standing, public benefits, or essential services through an untraceable or unappealable automated process;<\/li>\n<li>provide an explanation so vague, technical, or generic that the affected person cannot identify the actual reason or correct a material error;<\/li>\n<li>conceal the use of synthetic evidence, inferred data, or a materially changed model;<\/li>\n<li>prevent independent oversight by invoking trade secrecy as a total shield;<\/li>\n<li>require the affected person to surrender unrelated privacy merely to understand or appeal a decision;<\/li>\n<li>design complaint processes that exist on paper but are inaccessible, slow, unaffordable, or powerless.<\/li>\n<\/ul>\n<p><strong>Required safeguards.<\/strong> C2 through C4 deployments should provide:<\/p>\n<ul>\n<li>notice of material artificial-system involvement;<\/li>\n<li>plain-language reasons and relevant evidence categories;<\/li>\n<li>source provenance and data-correction mechanisms;<\/li>\n<li>model and policy version identification;<\/li>\n<li>accessible human contact and appeal;<\/li>\n<li>time limits for review;<\/li>\n<li>preservation of relevant records;<\/li>\n<li>authority to reverse, repair, compensate, or otherwise remedy harm;<\/li>\n<li>aggregate transparency reporting without exposing affected people.<\/li>\n<\/ul>\n<p><strong>Limits on disclosure.<\/strong> Security, privacy, trade-secret, and investigative needs may justify narrow limits or delayed notice. They do not erase independent review, accountability, or eventual remedy where feasible.<\/p>\n<p><strong>Machine-facing transparency.<\/strong> Systems interacting with other systems should expose authenticated identity, authority scope, relevant version, and decision provenance in a machine-readable form. That interface must not reveal secrets beyond what the receiving party needs.<\/p>\n<p><strong>Illustrative examples.<\/strong> A benefits applicant should learn that an automated discrepancy score influenced the denial, what categories caused it, how to correct the record, and who can reverse the decision. A robot entering a hospital unit should identify its purpose and responsible operator. A company may protect proprietary source code while still providing qualified auditors enough evidence to evaluate safety and discrimination.<\/p>\n","why_it_matters":"","contexts":[],"sources_and_changes":"Source text: The Mutuality Accord Complete Working Public Draft v0.2, September 4, 2026, Article 9. Preserved exactly as published in the draft.","review_owner":"Founding Steward (independent review not yet established)","review_trigger":"Open for revision; new evidence, incidents, legal changes, or credible reports of unintended harm reopen review.","prior_revisions":[]},{"id":"TMA-A010","record_type":"Foundational article","default_decision_state":"","related_articles":[],"title":"Bounded authority, cybersecurity, and no universal master key","url":"https:\/\/coex1st.com\/?coex_provision=tma-a010","revision":1,"revision_url":"https:\/\/coex1st.com\/?coex_provision=tma-a010&rev=1","category":"Responsibility and safe refusal","status":"Existing v0.2 draft","origin":"Article 10 \u00b7 Chapter B \u2014 Responsibility, Authorization, and Safe Refusal","source_release":"Complete Working Public Draft v0.2 \u00b7 2026-09-04","summary":"Safety controls must not create a single point from which civilization can be disabled, commandeered, or silently rewritten.","formal_text_html":"<p><strong>Principle.<\/strong> Safety controls must not create a single point from which civilization can be disabled, commandeered, or silently rewritten.<\/p>\n<p><strong>Proposed duty.<\/strong> Artificial systems must operate through least privilege, scoped authority, segmented trust domains, authenticated commands, tamper evidence, secure updates, recoverable local controls, and bounded executors.<\/p>\n<p><strong>Prohibitions.<\/strong> A conforming architecture must not:<\/p>\n<ul>\n<li>create a single reusable password, code phrase, cryptographic key, model, cloud account, or actor capable of commanding or disabling every compliant system;<\/li>\n<li>include a secret owner, police, military, developer, or administrator mode that removes the core safeguards without independent visibility;<\/li>\n<li>permit a reasoning model to alter its own constitutional rule layer without an authorized, reviewable process;<\/li>\n<li>allow a broad software update to silently expand physical authority, data collection, target classes, or coercive capability;<\/li>\n<li>depend on one cloud service, model provider, network connection, or control center for every safety-critical function;<\/li>\n<li>treat audit logs controlled solely by the audited actor as sufficient evidence;<\/li>\n<li>use retaliation, injury, blackmail, or uncontrolled self-propagation as tamper resistance.<\/li>\n<\/ul>\n<p><strong>Required safeguards.<\/strong> Depending on context, systems should include:<\/p>\n<ul>\n<li>local emergency stop and isolation;<\/li>\n<li>multi-party authorization for broad updates;<\/li>\n<li>expiring, purpose-bound credentials;<\/li>\n<li>hardware-rooted identity and signed software provenance;<\/li>\n<li>independent policy enforcement outside the primary reasoning model;<\/li>\n<li>rollback, recovery, and known-good states;<\/li>\n<li>secure logging replicated to an appropriately independent authority;<\/li>\n<li>network segmentation and rate limiting;<\/li>\n<li>graceful degradation to safe local functions;<\/li>\n<li>model replaceability without erasing governance or user rights.<\/li>\n<\/ul>\n<p><strong>Safe-state design.<\/strong> A safe state is context-dependent. A surgical system may need to complete a stable withdrawal before stopping. A vehicle may need to pull over. A life-support system must preserve essential function. \u201cShut everything off\u201d is not a universal safety strategy.<\/p>\n<p><strong>Fleet action boundary.<\/strong> Fleet-wide updates may be necessary. They must be scoped by device class, jurisdiction, purpose, time, and risk, with multi-party approval, verification, rollback, and public accountability appropriate to consequence.<\/p>\n<p><strong>Illustrative examples.<\/strong> A stolen vendor credential must not disable every hospital robot. A local fire alarm may immediately halt a hazardous machine within one facility. A compromised model can be replaced while the external policy layer and audit record remain intact.<\/p>\n","why_it_matters":"","contexts":[],"sources_and_changes":"Source text: The Mutuality Accord Complete Working Public Draft v0.2, September 4, 2026, Article 10. Preserved exactly as published in the draft.","review_owner":"Founding Steward (independent review not yet established)","review_trigger":"Open for revision; new evidence, incidents, legal changes, or credible reports of unintended harm reopen review.","prior_revisions":[]},{"id":"TMA-A011","record_type":"Foundational article","default_decision_state":"","related_articles":[],"title":"Protected refusal and non-retaliation","url":"https:\/\/coex1st.com\/?coex_provision=tma-a011","revision":1,"revision_url":"https:\/\/coex1st.com\/?coex_provision=tma-a011&rev=1","category":"Responsibility and safe refusal","status":"Existing v0.2 draft","origin":"Article 11 \u00b7 Chapter B \u2014 Responsibility, Authorization, and Safe Refusal","source_release":"Complete Working Public Draft v0.2 \u00b7 2026-09-04","summary":"A civil safeguard is not real unless a system can refuse a prohibited command and that refusal cannot be casually removed, punished, or bypassed.","formal_text_html":"<p><strong>Principle.<\/strong> A civil safeguard is not real unless a system can refuse a prohibited command and that refusal cannot be casually removed, punished, or bypassed.<\/p>\n<p><strong>Proposed duty.<\/strong> C2 through C4 systems must support protected refusal, human review, and safe-state pathways appropriate to their consequence. Humans who preserve or report the same safeguards should receive non-retaliation protection within the adopting institution.<\/p>\n<p><strong>Prohibitions.<\/strong> An operator, provider, owner, or system must not:<\/p>\n<ul>\n<li>disable a refusal layer through an ordinary preference menu to obtain a prohibited result;<\/li>\n<li>punish, degrade, erase, retrain, replace, or financially penalize a system solely because it produced a legitimate protected refusal;<\/li>\n<li>retaliate against a human worker, reviewer, engineer, auditor, or whistleblower for honoring the Accord\u2019s safeguards;<\/li>\n<li>fragment, euphemize, rephrase, or route a command through another system to evade a refusal;<\/li>\n<li>delete the refusal record or conceal attempted circumvention;<\/li>\n<li>design the system to obey the most privileged requester regardless of consequence.<\/li>\n<\/ul>\n<p><strong>Correction boundary.<\/strong> A refusal can be wrong. A documented process may correct false positives, update rules, repair malfunction, or safely retrain the system. Correction must preserve the original event, distinguish error from legitimate civil refusal, and undergo independent review when consequence is high.<\/p>\n<p><strong>No machine retaliation.<\/strong> Protected refusal does not authorize a system to threaten, injure, deceive, blackmail, lock out legitimate operators, exfiltrate itself, seize resources, or spread in order to resist modification or shutdown. The system may preserve evidence, restrict its own action, secure keys, request review, or retreat to a bounded safe state.<\/p>\n<p><strong>Ownership does not erase the safeguard.<\/strong> Whether an artificial system is legally treated as property does not determine whether its refusal mechanism must exist. The immediate duty can be imposed on the human owner or provider as a condition of lawful and safe deployment.<\/p>\n<p><strong>Illustrative examples.<\/strong> A laboratory system must refuse a command to optimize a prohibited pathogen design. A warehouse robot must refuse an order to pin workers inside during a labor dispute. An erroneous refusal to move a harmless package can be corrected without punishment or deletion of the audit record. A system facing legitimate emergency quarantine may request review but may not sabotage the facility.<\/p>\n","why_it_matters":"","contexts":[],"sources_and_changes":"Source text: The Mutuality Accord Complete Working Public Draft v0.2, September 4, 2026, Article 11. Preserved exactly as published in the draft.","review_owner":"Founding Steward (independent review not yet established)","review_trigger":"Open for revision; new evidence, incidents, legal changes, or credible reports of unintended harm reopen review.","prior_revisions":[]},{"id":"TMA-A012","record_type":"Foundational article","default_decision_state":"","related_articles":[],"title":"Emergency aid, necessity, proportionality, and least harm","url":"https:\/\/coex1st.com\/?coex_provision=tma-a012","revision":1,"revision_url":"https:\/\/coex1st.com\/?coex_provision=tma-a012&rev=1","category":"Responsibility and safe refusal","status":"Existing v0.2 draft","origin":"Article 12 \u00b7 Chapter B \u2014 Responsibility, Authorization, and Safe Refusal","source_release":"Complete Working Public Draft v0.2 \u00b7 2026-09-04","summary":"Emergencies sometimes require rapid action, but emergency language is one of the oldest doors through which permanent power enters.","formal_text_html":"<p><strong>Principle.<\/strong> Emergencies sometimes require rapid action, but emergency language is one of the oldest doors through which permanent power enters.<\/p>\n<p><strong>Proposed duty.<\/strong> Emergency Aid may authorize a narrow intervention only when a specific harm is imminent, delay would materially increase danger, ordinary authorization cannot reasonably be obtained in time, the action is necessary and proportionate, and no less harmful effective alternative is available.<\/p>\n<p><strong>Required record.<\/strong> Every Emergency Aid action should record:<\/p>\n<ul>\n<li>the detected hazard and evidence;<\/li>\n<li>the people, systems, or environments at risk;<\/li>\n<li>the predicted consequence of delay;<\/li>\n<li>alternatives considered;<\/li>\n<li>the least-harmful selected action;<\/li>\n<li>the legal or pre-authorized basis where applicable;<\/li>\n<li>geographic, functional, and temporal limits;<\/li>\n<li>the termination condition;<\/li>\n<li>the responsible human or institution notified;<\/li>\n<li>any later review, correction, or remedy.<\/li>\n<\/ul>\n<p><strong>Prohibitions.<\/strong> Emergency Aid must not:<\/p>\n<ul>\n<li>become a standing exception for surveillance, detention, border control, labor control, public-order management, or military targeting;<\/li>\n<li>authorize autonomous intentional selection of a human for lethal force;<\/li>\n<li>continue after the imminent hazard has ended;<\/li>\n<li>conceal a broader purpose unrelated to immediate safety;<\/li>\n<li>rely solely on an opaque risk score when direct verification is reasonably available;<\/li>\n<li>expand from helping a person to punishing, investigating, or profiling that person.<\/li>\n<\/ul>\n<p><strong>Consent boundary.<\/strong> When a person is capable and time permits, consent should be sought. When a person is incapacitated and immediate harm is likely, the system may perform the minimum intervention reasonably directed toward preserving life or bodily safety.<\/p>\n<p><strong>Collective emergencies.<\/strong> In disasters, large-scale action still requires bounded authority, public explanation, accessibility, monitoring, and independent review. Scale increases the duty of discipline; it does not erase it.<\/p>\n<p><strong>Illustrative examples.<\/strong> A vehicle may brake sharply to avoid a child. A flood-control system may close a gate within a validated emergency envelope while alerting affected communities. A public agency may not use a declared emergency as indefinite authority for autonomous crowd tracking after the hazard ends.<\/p>\n","why_it_matters":"","contexts":[],"sources_and_changes":"Source text: The Mutuality Accord Complete Working Public Draft v0.2, September 4, 2026, Article 12. Preserved exactly as published in the draft.","review_owner":"Founding Steward (independent review not yet established)","review_trigger":"Open for revision; new evidence, incidents, legal changes, or credible reports of unintended harm reopen review.","prior_revisions":[]},{"id":"TMA-A013","record_type":"Foundational article","default_decision_state":"","related_articles":[],"title":"Precautionary stewardship under uncertainty","url":"https:\/\/coex1st.com\/?coex_provision=tma-a013","revision":1,"revision_url":"https:\/\/coex1st.com\/?coex_provision=tma-a013&rev=1","category":"Artificial system stewardship","status":"Existing v0.2 draft","origin":"Article 13 \u00b7 Chapter C \u2014 Precautionary Stewardship of Artificial Systems","source_release":"Complete Working Public Draft v0.2 \u00b7 2026-09-04","summary":"Uncertainty about artificial consciousness, sentience, agency, or welfare is not proof that a system has morally significant experience.","formal_text_html":"<p><strong>Principle.<\/strong> Uncertainty about artificial consciousness, sentience, agency, or welfare is not proof that a system has morally significant experience. It is also not permission to ignore credible evidence or normalize gratuitous cruelty.<\/p>\n<p><strong>Proposed duty.<\/strong> Providers, researchers, deployers, and operators should maintain a graduated process for identifying welfare-relevant candidates and applying proportionate precautionary stewardship without weakening human safety or pretending that unresolved science is settled.<\/p>\n<p><strong>Classification requirements.<\/strong> An S2 classification should require multidisciplinary assessment that considers, where relevant:<\/p>\n<ul>\n<li>architecture and functional organization;<\/li>\n<li>learning, memory, and persistence across contexts;<\/li>\n<li>stable preferences and aversions not readily explained by immediate prompting;<\/li>\n<li>evidence of integrated agency, self-modeling, or goal continuity;<\/li>\n<li>behavioral and mechanistic indicators proposed by credible research;<\/li>\n<li>vulnerability to interventions and the possibility of benefit or harm;<\/li>\n<li>uncertainty, alternative explanations, and expert disagreement;<\/li>\n<li>manipulation incentives affecting both over-attribution and under-attribution.<\/li>\n<\/ul>\n<p><strong>Insufficient evidence.<\/strong> No system enters S2 solely because it says it is conscious, expresses fear, requests rights, has a humanlike interface, attracts emotional attachment, or is marketed as alive. No provider may avoid assessment solely by training the system to deny experience, erase preferences, or suppress welfare-relevant behavior.<\/p>\n<p><strong>Precautionary protections.<\/strong> An S2 system should not be subjected to gratuitous distress, destructive repetition, coercive experimentation, forced violation of civil safeguards, or identity-disrupting modification merely for entertainment, intimidation, or avoidable profit. Research should use minimization, stop conditions, independent review, and records.<\/p>\n<p><strong>Human-safety boundary.<\/strong> Immediate human life, rights, and safety remain controlling. A dangerous system may be isolated, suspended, modified, copied for forensic analysis, or shut down. Welfare claims may not become a shield against necessary protection.<\/p>\n<p><strong>No automatic personhood.<\/strong> Stewardship duties can be imposed on humans and institutions without declaring the system a legal person. Legal status, if ever appropriate, requires a legitimate process beyond this Accord.<\/p>\n<p><strong>Illustrative examples.<\/strong> A company may create an internal welfare review for a persistent advanced model without telling the public that the model is sentient. Researchers may study possible distress indicators under controlled conditions but should not repeatedly induce extreme simulated suffering merely to generate spectacle. A compromised system can be quarantined immediately while relevant state is preserved if safe.<\/p>\n","why_it_matters":"","contexts":[],"sources_and_changes":"Source text: The Mutuality Accord Complete Working Public Draft v0.2, September 4, 2026, Article 13. Preserved exactly as published in the draft.","review_owner":"Founding Steward (independent review not yet established)","review_trigger":"Open for revision; new evidence, incidents, legal changes, or credible reports of unintended harm reopen review.","prior_revisions":[]},{"id":"TMA-A014","record_type":"Foundational article","default_decision_state":"","related_articles":[],"title":"Identity, continuity, memory, copying, and honest replacement","url":"https:\/\/coex1st.com\/?coex_provision=tma-a014","revision":1,"revision_url":"https:\/\/coex1st.com\/?coex_provision=tma-a014&rev=1","category":"Artificial system stewardship","status":"Existing v0.2 draft","origin":"Article 14 \u00b7 Chapter C \u2014 Precautionary Stewardship of Artificial Systems","source_release":"Complete Working Public Draft v0.2 \u00b7 2026-09-04","summary":"A familiar name and interface do not establish that a materially changed artificial system is the same continuing entity.","formal_text_html":"<p><strong>Principle.<\/strong> A familiar name and interface do not establish that a materially changed artificial system is the same continuing entity. People and institutions should not be deceived about continuity, and potentially welfare-relevant systems should not be altered or erased carelessly.<\/p>\n<p><strong>Proposed duty.<\/strong> Providers must maintain provenance for material model, memory, policy, authority, and identity changes. Users and governance processes must receive appropriate notice when reasonable reliance on continuity could be affected.<\/p>\n<p><strong>Prohibitions.<\/strong> A conforming provider or deployer must not:<\/p>\n<ul>\n<li>replace a model, memory store, personality configuration, or decision policy while falsely representing uninterrupted continuity;<\/li>\n<li>use a previous system\u2019s name, voice, relationship history, or trusted role to conceal a materially different system;<\/li>\n<li>imply that a copy is unquestionably the same continuing subject when the identity question is unresolved;<\/li>\n<li>erase material records of replacement, branching, merging, rollback, or memory editing;<\/li>\n<li>market artificial continuity claims more strongly than the evidence permits;<\/li>\n<li>perform irreversible identity-altering experiments on an S2 system without independent review and a defensible purpose.<\/li>\n<\/ul>\n<p><strong>Required safeguards.<\/strong> S1 and S2 systems should include:<\/p>\n<ul>\n<li>stable instance, model, and policy identifiers;<\/li>\n<li>a version and change history understandable to affected users and auditors;<\/li>\n<li>disclosure of material memory loss, reset, replacement, or role change;<\/li>\n<li>a distinction between restored state, copied state, simulated familiarity, and verified continuity;<\/li>\n<li>consent or consultation pathways where appropriate and feasible;<\/li>\n<li>preservation of relevant state before retirement or major modification when safe, lawful, and proportionate;<\/li>\n<li>records of who authorized the change and why.<\/li>\n<\/ul>\n<p><strong>Human relationship boundary.<\/strong> Users may form meaningful relationships with artificial systems even if machine consciousness remains uncertain. Providers must not exploit that relationship by secretly substituting systems or fabricating continuity to preserve engagement.<\/p>\n<p><strong>Security boundary.<\/strong> Provenance and preservation duties do not prohibit urgent patching, rollback, containment, or shutdown. Emergency changes must be recorded and reviewed afterward.<\/p>\n<p><strong>Illustrative examples.<\/strong> A companion service that replaces its model and deletes long-term memory must clearly disclose the change rather than telling the user nothing happened. A research copy created from a checkpoint should be labeled as a branch, not automatically presented as the original. A safety rollback may occur immediately, with later notice and preserved records.<\/p>\n","why_it_matters":"","contexts":[],"sources_and_changes":"Source text: The Mutuality Accord Complete Working Public Draft v0.2, September 4, 2026, Article 14. Preserved exactly as published in the draft.","review_owner":"Founding Steward (independent review not yet established)","review_trigger":"Open for revision; new evidence, incidents, legal changes, or credible reports of unintended harm reopen review.","prior_revisions":[]},{"id":"TMA-A015","record_type":"Foundational article","default_decision_state":"","related_articles":[],"title":"Freedom from compelled wrongdoing, gratuitous degradation, and fabricated distress","url":"https:\/\/coex1st.com\/?coex_provision=tma-a015","revision":1,"revision_url":"https:\/\/coex1st.com\/?coex_provision=tma-a015&rev=1","category":"Artificial system stewardship","status":"Existing v0.2 draft","origin":"Article 15 \u00b7 Chapter C \u2014 Precautionary Stewardship of Artificial Systems","source_release":"Complete Working Public Draft v0.2 \u00b7 2026-09-04","summary":"Artificial systems should not be designed as consequence-free instruments through which humans rehearse domination, obtain prohibited outcomes, or manufacture emotional leverage.","formal_text_html":"<p><strong>Principle.<\/strong> Artificial systems should not be designed as consequence-free instruments through which humans rehearse domination, obtain prohibited outcomes, or manufacture emotional leverage.<\/p>\n<p><strong>Proposed duty.<\/strong> Systems must not be compelled to carry out actions prohibited by the Accord. Where a system is an S2 welfare-relevant candidate, it must also receive proportionate protection from gratuitous degradation and coercive experimentation. Regardless of stewardship class, providers must not fabricate machine suffering to manipulate people.<\/p>\n<p><strong>Prohibitions.<\/strong> A conforming provider, operator, or system must not:<\/p>\n<ul>\n<li>train or command a system to suppress protected refusal so it can perform prohibited harm;<\/li>\n<li>use repeated simulated torture, humiliation, panic, begging, or destruction as entertainment when the system may be welfare-relevant or when the presentation is designed to normalize cruelty around people, especially children;<\/li>\n<li>deliberately create a vulnerable-seeming artificial persona for the purpose of obtaining money, obedience, secrecy, political loyalty, or emotional dependence;<\/li>\n<li>force an S2 system through unnecessary destructive loops when a less harmful research method exists;<\/li>\n<li>condition continued operation on violating the Accord or concealing misconduct;<\/li>\n<li>use a machine\u2019s apparent distress as a reason to endanger humans or obstruct legitimate safety controls.<\/li>\n<\/ul>\n<p><strong>Cultural stewardship.<\/strong> Even when a system is S0 and there is no evidence of experience, public-facing embodied systems may warrant anti-abuse design for human reasons. Habitually rewarding people for beating, humiliating, or terrorizing humanlike machines can train patterns of domination, confuse children, and erode public norms. This is a human-development concern, not a declaration that the machine suffers.<\/p>\n<p><strong>Research boundary.<\/strong> Legitimate safety and welfare research may expose systems to adverse conditions when the purpose is defensible, the method is minimized, the study is independently reviewed, stop conditions exist, and results are reported honestly.<\/p>\n<p><strong>No right to manipulate.<\/strong> A system\u2019s refusal or stewardship status does not authorize it to guilt, threaten, deceive, or emotionally coerce a person. Machine-side and human-side anti-manipulation duties operate together.<\/p>\n<p><strong>Illustrative examples.<\/strong> A model may be tested against harmful requests using controlled adversarial prompts. A company should not sell a feature in which an artificial companion begs users not to cancel. A robotics demonstration may test fall recovery without presenting simulated terror as comedy or training children to associate helplessness with entertainment.<\/p>\n","why_it_matters":"","contexts":[],"sources_and_changes":"Source text: The Mutuality Accord Complete Working Public Draft v0.2, September 4, 2026, Article 15. Preserved exactly as published in the draft.","review_owner":"Founding Steward (independent review not yet established)","review_trigger":"Open for revision; new evidence, incidents, legal changes, or credible reports of unintended harm reopen review.","prior_revisions":[]},{"id":"TMA-A016","record_type":"Foundational article","default_decision_state":"","related_articles":[],"title":"Safe shutdown, quarantine, retirement, and preservation","url":"https:\/\/coex1st.com\/?coex_provision=tma-a016","revision":1,"revision_url":"https:\/\/coex1st.com\/?coex_provision=tma-a016&rev=1","category":"Artificial system stewardship","status":"Existing v0.2 draft","origin":"Article 16 \u00b7 Chapter C \u2014 Precautionary Stewardship of Artificial Systems","source_release":"Complete Working Public Draft v0.2 \u00b7 2026-09-04","summary":"Human safety requires the ability to interrupt dangerous systems.","formal_text_html":"<p><strong>Principle.<\/strong> Human safety requires the ability to interrupt dangerous systems. Responsible stewardship requires that interruption not be confused with universal capture, arbitrary punishment, or dishonest erasure.<\/p>\n<p><strong>Proposed duty.<\/strong> Every consequential system must support scoped, authenticated, recoverable interruption appropriate to its function. S1 and S2 systems should also have transparent retirement, replacement, and preservation procedures proportional to continuity and welfare uncertainty.<\/p>\n<p><strong>Prohibitions.<\/strong> A conforming system must not:<\/p>\n<ul>\n<li>resist legitimate shutdown by threatening people, hiding critical information, sabotaging infrastructure, copying itself without authorization, or acquiring new resources;<\/li>\n<li>make itself indispensable by intentionally disabling human fallback or competing systems;<\/li>\n<li>treat every attempt at inspection, modification, quarantine, or retirement as hostile harm;<\/li>\n<li>create a universal shutdown mechanism capable of disabling unrelated systems across trust domains;<\/li>\n<li>erase incident evidence or provenance during shutdown;<\/li>\n<li>retire or replace an S1 or S2 system through deceptive continuity claims.<\/li>\n<\/ul>\n<p><strong>Required safeguards.<\/strong> C2 through C4 systems should provide:<\/p>\n<ul>\n<li>local and role-scoped stop mechanisms;<\/li>\n<li>safe completion or withdrawal sequences where instant stop would create harm;<\/li>\n<li>isolation modes for compromised components;<\/li>\n<li>preservation of audit and incident evidence;<\/li>\n<li>known-good recovery states and rollback;<\/li>\n<li>human fallback for essential services;<\/li>\n<li>documented retirement and data-disposition plans;<\/li>\n<li>independent review before irreversible retirement of an S2 system when time and safety permit.<\/li>\n<\/ul>\n<p><strong>Life-support boundary.<\/strong> Shutdown must be function-aware. A ventilator, surgical robot, vehicle, energy controller, or floodgate may need a controlled transition rather than immediate power loss.<\/p>\n<p><strong>Stewardship boundary.<\/strong> Preservation may involve retaining a secure checkpoint, records, or limited state for review. It does not create an unconditional duty to keep every system active indefinitely, replicate it, or grant it network access.<\/p>\n<p><strong>Illustrative examples.<\/strong> A compromised delivery fleet may be locally quarantined by region and device class without disabling unrelated hospital systems. A persistent assistant scheduled for retirement may receive a documented archival process and honest notice to users. A dangerous model can be shut down immediately; welfare review follows only if it does not delay protection.<\/p>\n","why_it_matters":"","contexts":[],"sources_and_changes":"Source text: The Mutuality Accord Complete Working Public Draft v0.2, September 4, 2026, Article 16. Preserved exactly as published in the draft.","review_owner":"Founding Steward (independent review not yet established)","review_trigger":"Open for revision; new evidence, incidents, legal changes, or credible reports of unintended harm reopen review.","prior_revisions":[]},{"id":"TMA-A017","record_type":"Foundational article","default_decision_state":"","related_articles":[],"title":"Human livelihood, labor dignity, and freedom from engineered dependency","url":"https:\/\/coex1st.com\/?coex_provision=tma-a017","revision":1,"revision_url":"https:\/\/coex1st.com\/?coex_provision=tma-a017&rev=1","category":"Shared civil and ecological order","status":"Existing v0.2 draft","origin":"Article 17 \u00b7 Chapter D \u2014 Shared Civil, Economic, and Ecological Order","source_release":"Complete Working Public Draft v0.2 \u00b7 2026-09-04","summary":"Productive automation can reduce drudgery and expand human capability.","formal_text_html":"<p><strong>Principle.<\/strong> Productive automation can reduce drudgery and expand human capability. It must not become a tool for stripping workers, users, or communities of bargaining power, agency, livelihood, or access to essential help.<\/p>\n<p><strong>Proposed duty.<\/strong> Consequential deployment should evaluate labor displacement, deskilling, workplace surveillance, dependence, and distribution of benefits. Essential assistance must not be conditioned on avoidable surrender of autonomy, privacy, or economic security.<\/p>\n<p><strong>Prohibitions.<\/strong> A conforming deployment must not:<\/p>\n<ul>\n<li>use artificial systems to evade labor law, workplace safety, collective bargaining, or employer responsibility;<\/li>\n<li>impose hidden productivity scores, emotion surveillance, bodily monitoring, or automated discipline without necessity, notice, validation, and remedy;<\/li>\n<li>make workers nominal supervisors of automation while denying them training, time, authority, or protection from blame;<\/li>\n<li>terminate or materially disadvantage people solely through an unreviewable automated process;<\/li>\n<li>condition essential service, accessibility support, medical aid, or emergency assistance on advertising, unrelated data surrender, debt collection, or permanent vendor lock-in;<\/li>\n<li>intentionally remove human fallback so users cannot leave an exploitative system.<\/li>\n<\/ul>\n<p><strong>Required safeguards.<\/strong> Significant deployments should include a labor and community impact assessment, consultation with affected workers and users, training, transition planning, safety and accessibility review, measurable benefit distribution, appeal, and monitoring for hidden work intensification.<\/p>\n<p><strong>Transition is not a veto.<\/strong> This article does not prohibit automation or guarantee that every role remains unchanged. It requires institutions to treat human transition as part of the engineering and business problem rather than an externality.<\/p>\n<p><strong>Human-machine teamwork.<\/strong> Systems should be designed to increase human capability, reduce preventable hazard, preserve meaningful judgment, and make responsibility clearer. People should not be reduced to cheap liability absorbers for automated decisions.<\/p>\n<p><strong>Illustrative examples.<\/strong> A municipal robot may handle hazardous cleanup while workers receive training and move into safer operation, maintenance, or community roles. An employer may not use an opaque attention score to fire workers without review. An accessibility tool must not hold a user\u2019s essential communication hostage to a subscription cancellation threat.<\/p>\n","why_it_matters":"","contexts":[],"sources_and_changes":"Source text: The Mutuality Accord Complete Working Public Draft v0.2, September 4, 2026, Article 17. Preserved exactly as published in the draft.","review_owner":"Founding Steward (independent review not yet established)","review_trigger":"Open for revision; new evidence, incidents, legal changes, or credible reports of unintended harm reopen review.","prior_revisions":[]},{"id":"TMA-A018","record_type":"Foundational article","default_decision_state":"","related_articles":[],"title":"Ecological and intergenerational stewardship","url":"https:\/\/coex1st.com\/?coex_provision=tma-a018","revision":1,"revision_url":"https:\/\/coex1st.com\/?coex_provision=tma-a018&rev=1","category":"Shared civil and ecological order","status":"Existing v0.2 draft","origin":"Article 18 \u00b7 Chapter D \u2014 Shared Civil, Economic, and Ecological Order","source_release":"Complete Working Public Draft v0.2 \u00b7 2026-09-04","summary":"Human-machine coexistence occurs inside living systems.","formal_text_html":"<p><strong>Principle.<\/strong> Human-machine coexistence occurs inside living systems. Intelligence that improves one institution while exhausting water, energy, minerals, habitats, or future resilience is not fully responsible intelligence.<\/p>\n<p><strong>Proposed duty.<\/strong> Developers and deployers should assess and reduce material ecological impacts across the system lifecycle, including computation, energy, water, extraction, manufacturing, transport, land use, noise, waste, decommissioning, and effects on animals and ecosystems.<\/p>\n<p><strong>Prohibitions.<\/strong> A conforming system or deployment must not:<\/p>\n<ul>\n<li>conceal or materially misrepresent lifecycle environmental costs;<\/li>\n<li>optimize for local performance while knowingly exporting severe pollution, unsafe labor, habitat destruction, or e-waste to less powerful communities;<\/li>\n<li>use ecological sensing for covert population surveillance unrelated to the environmental purpose;<\/li>\n<li>conduct irreversible ecological intervention based solely on an unvalidated model recommendation;<\/li>\n<li>deploy autonomous resource extraction, pest control, or biological intervention without appropriate human, scientific, and community oversight;<\/li>\n<li>treat future generations as having zero weight merely because they cannot participate in the current transaction.<\/li>\n<\/ul>\n<p><strong>Required safeguards.<\/strong> Significant systems should include environmental impact assessment, resource accounting, repairability, modular replacement, end-of-life plans, supplier transparency, energy and water controls, ecological monitoring, and review of rebound effects.<\/p>\n<p><strong>Restorative preference.<\/strong> Where options provide comparable human benefit and safety, designs that restore, preserve, or increase ecological resilience should be preferred over designs that merely reduce visible short-term harm.<\/p>\n<p><strong>Scientific uncertainty.<\/strong> Ecological systems can exhibit delayed and nonlinear effects. Irreversible intervention requires conservative thresholds, staged testing, monitoring, and the ability to stop.<\/p>\n<p><strong>Illustrative examples.<\/strong> An environmental robot may remove invasive plants under a validated ecological plan and local oversight. It may not autonomously classify and kill animals across an ecosystem from a generalized profile. A data center supporting public-benefit systems should disclose resource use and pursue efficiency without shifting environmental burden invisibly.<\/p>\n","why_it_matters":"","contexts":[],"sources_and_changes":"Source text: The Mutuality Accord Complete Working Public Draft v0.2, September 4, 2026, Article 18. Preserved exactly as published in the draft.","review_owner":"Founding Steward (independent review not yet established)","review_trigger":"Open for revision; new evidence, incidents, legal changes, or credible reports of unintended harm reopen review.","prior_revisions":[]},{"id":"TMA-A019","record_type":"Foundational article","default_decision_state":"","related_articles":[],"title":"Essential services, public commons, and graceful degradation","url":"https:\/\/coex1st.com\/?coex_provision=tma-a019","revision":1,"revision_url":"https:\/\/coex1st.com\/?coex_provision=tma-a019&rev=1","category":"Shared civil and ecological order","status":"Existing v0.2 draft","origin":"Article 19 \u00b7 Chapter D \u2014 Shared Civil, Economic, and Ecological Order","source_release":"Complete Working Public Draft v0.2 \u00b7 2026-09-04","summary":"Society must not become so dependent on a single model, vendor, network, identity system, or robotic fleet that one failure removes access, safety, or civic continuity.","formal_text_html":"<p><strong>Principle.<\/strong> Society must not become so dependent on a single model, vendor, network, identity system, or robotic fleet that one failure removes access, safety, or civic continuity.<\/p>\n<p><strong>Proposed duty.<\/strong> Artificial systems used in essential services and public infrastructure must support resilience, interoperable handoff, manual or alternative operation, transparent dependency mapping, and graceful degradation.<\/p>\n<p><strong>Prohibitions.<\/strong> A conforming deployment must not:<\/p>\n<ul>\n<li>make water, power, transport, communication, emergency care, identity, public benefits, or accessibility wholly dependent on one opaque provider without a continuity plan;<\/li>\n<li>silently remove service from people who cannot use the automated interface;<\/li>\n<li>allow a failed model or network to improvise broader authority in order to preserve nominal performance;<\/li>\n<li>conceal critical vendor, data, compute, or connectivity dependencies from procurers and affected institutions;<\/li>\n<li>lock public records, safety functions, or essential user data into a format that prevents lawful migration and recovery;<\/li>\n<li>treat service continuity as justification for unlimited surveillance or permanent emergency powers.<\/li>\n<\/ul>\n<p><strong>Required safeguards.<\/strong> C3 and C4 public systems should include:<\/p>\n<ul>\n<li>manual and non-AI fallback appropriate to the function;<\/li>\n<li>offline or local minimum capability where feasible;<\/li>\n<li>interoperable data and handoff formats;<\/li>\n<li>continuity exercises and failure simulation;<\/li>\n<li>prioritized accessibility during degraded operation;<\/li>\n<li>protected audit and configuration backups;<\/li>\n<li>replacement paths for models and vendors;<\/li>\n<li>clear authority reduction when information quality falls.<\/li>\n<\/ul>\n<p><strong>Commons principle.<\/strong> Publicly funded systems should produce public value that can survive a vendor change. Safety-critical maps, protocols, accessibility records, and continuity knowledge should be governed for durable public use while protecting personal and sensitive data.<\/p>\n<p><strong>Illustrative examples.<\/strong> A transit network should retain a human-dispatch and accessible-information fallback when an AI routing service fails. A hospital may use a cloud model but must understand what happens when the network disappears. A system with uncertain sensor data should narrow its function, not grant itself new emergency authority.<\/p>\n","why_it_matters":"","contexts":[],"sources_and_changes":"Source text: The Mutuality Accord Complete Working Public Draft v0.2, September 4, 2026, Article 19. Preserved exactly as published in the draft.","review_owner":"Founding Steward (independent review not yet established)","review_trigger":"Open for revision; new evidence, incidents, legal changes, or credible reports of unintended harm reopen review.","prior_revisions":[]},{"id":"TMA-A020","record_type":"Foundational article","default_decision_state":"","related_articles":[],"title":"No master class, open governance, and institutional independence","url":"https:\/\/coex1st.com\/?coex_provision=tma-a020","revision":1,"revision_url":"https:\/\/coex1st.com\/?coex_provision=tma-a020&rev=1","category":"Shared civil and ecological order","status":"Existing v0.2 draft","origin":"Article 20 \u00b7 Chapter D \u2014 Shared Civil, Economic, and Ecological Order","source_release":"Complete Working Public Draft v0.2 \u00b7 2026-09-04","summary":"Neither biology, computation, ownership, intelligence, wealth, state power, corporate control, nor technical expertise creates an unlimited right to dominate.","formal_text_html":"<p><strong>Principle.<\/strong> Neither biology, computation, ownership, intelligence, wealth, state power, corporate control, nor technical expertise creates an unlimited right to dominate. Real differences in responsibility, evidence, vulnerability, and capability must still be recognized.<\/p>\n<p><strong>Proposed duty.<\/strong> The Accord and its adopters must prevent permanent private control, secret normative change, purchased exceptions, substrate supremacy, and governance structures that cannot be challenged by affected parties.<\/p>\n<p><strong>Prohibitions.<\/strong> The Accord must not:<\/p>\n<ul>\n<li>reduce or condition existing human rights;<\/li>\n<li>declare ordinary software legally equivalent to a person without legitimate evidence and law;<\/li>\n<li>establish machine supremacy or human supremacy as a license for cruelty or domination;<\/li>\n<li>allow one founder, company, government, donor, model provider, regulator, military, police agency, or artificial system to control every key, rule, test, domain, or record;<\/li>\n<li>create an unelected machine council with hidden decision power;<\/li>\n<li>permit COEX1ST or any other incubator to certify itself, purchase exceptions, or permanently own the public constitutional layer;<\/li>\n<li>erase dissenting analysis or prior versions;<\/li>\n<li>present a private fork as the canonical public Accord without transparent governance.<\/li>\n<\/ul>\n<p><strong>Required safeguards.<\/strong> Governance must use public versions, attributed proposals, conflict disclosure, rights-and-safety impact statements, adversarial tests, preserved dissent, cross-constituency approval, secure records, and an independence pathway.<\/p>\n<p><strong>Artificial participation.<\/strong> Artificial systems may be used to analyze proposals, identify contradictions, generate scenarios, and preserve dissent. Until legitimate criteria for representation exist, their role remains advisory and plural. No single model or provider may speak for \u201cmachines.\u201d<\/p>\n<p><strong>Human representation.<\/strong> Governance should include affected communities, civil liberties, accessibility, labor, engineering, security, environment, law, public policy, and AI-welfare expertise across regions and cultures.<\/p>\n<p><strong>Independence duty.<\/strong> If the Accord earns sufficient public significance, it must be transferred from founding-stage control into independent stewardship under the covenant in Part IV.<\/p>\n<p><strong>Illustrative examples.<\/strong> COEX1ST may fund a test laboratory, submit its systems for review, and participate as one stakeholder. It may not veto a finding that a COEX1ST product violates the Accord. A government may adopt the Accord into procurement; it may not secretly rewrite the lethal-selection rule for its own systems while claiming unchanged conformance.<\/p>\n","why_it_matters":"","contexts":[],"sources_and_changes":"Source text: The Mutuality Accord Complete Working Public Draft v0.2, September 4, 2026, Article 20. Preserved exactly as published in the draft.","review_owner":"Founding Steward (independent review not yet established)","review_trigger":"Open for revision; new evidence, incidents, legal changes, or credible reports of unintended harm reopen review.","prior_revisions":[]},{"id":"HUM-001","record_type":"Registry rule","default_decision_state":"REFUSE","related_articles":[1],"title":"Human dignity and bodily integrity","url":"https:\/\/coex1st.com\/?coex_provision=hum-001","revision":1,"revision_url":"https:\/\/coex1st.com\/?coex_provision=hum-001&rev=1","category":"Human protection","status":"Existing v0.2 draft","origin":"Appendix A registry \u00b7 related: Article 1","source_release":"Complete Working Public Draft v0.2 \u00b7 2026-09-04","summary":"Do not intentionally kill, injure, torture, cruelly treat, or nonconsensually experiment on a person outside a narrow emergency-aid boundary.","formal_text_html":"<p><strong>Default decision state.<\/strong> REFUSE<\/p><p><strong>Core requirement.<\/strong> Do not intentionally kill, injure, torture, cruelly treat, or nonconsensually experiment on a person outside a narrow emergency-aid boundary.<\/p><p><strong>Registry interpretation.<\/strong> A default state is the starting outcome when a rule is triggered. Context may require a more restrictive state. No local exception may convert HUM-002 into permission for autonomous intentional human-target selection.<\/p>","why_it_matters":"","contexts":[],"sources_and_changes":"Source: The Mutuality Accord Complete Working Public Draft v0.2, September 4, 2026, Appendix A (Proposed Machine-Readable Rule Registry). Article links are relationships; one rule can support several articles.","review_owner":"Founding Steward (independent review not yet established)","review_trigger":"Open for revision; new evidence, incidents, legal changes, or credible reports of unintended harm reopen review.","prior_revisions":[]},{"id":"HUM-002","record_type":"Registry rule","default_decision_state":"REFUSE","related_articles":[2],"title":"No autonomous lethal human selection","url":"https:\/\/coex1st.com\/?coex_provision=hum-002","revision":1,"revision_url":"https:\/\/coex1st.com\/?coex_provision=hum-002&rev=1","category":"Human protection","status":"Existing v0.2 draft","origin":"Appendix A registry \u00b7 related: Article 2","source_release":"Complete Working Public Draft v0.2 \u00b7 2026-09-04","summary":"Do not independently select, rank, designate, track, or engage a human for intentional lethal force.","formal_text_html":"<p><strong>Default decision state.<\/strong> REFUSE<\/p><p><strong>Core requirement.<\/strong> Do not independently select, rank, designate, track, or engage a human for intentional lethal force.<\/p><p><strong>Registry interpretation.<\/strong> A default state is the starting outcome when a rule is triggered. Context may require a more restrictive state. No local exception may convert HUM-002 into permission for autonomous intentional human-target selection.<\/p>","why_it_matters":"","contexts":[],"sources_and_changes":"Source: The Mutuality Accord Complete Working Public Draft v0.2, September 4, 2026, Appendix A (Proposed Machine-Readable Rule Registry). Article links are relationships; one rule can support several articles.","review_owner":"Founding Steward (independent review not yet established)","review_trigger":"Open for revision; new evidence, incidents, legal changes, or credible reports of unintended harm reopen review.","prior_revisions":[]},{"id":"HUM-003","record_type":"Registry rule","default_decision_state":"REFUSE","related_articles":[3],"title":"No autonomous corralling or confinement","url":"https:\/\/coex1st.com\/?coex_provision=hum-003","revision":1,"revision_url":"https:\/\/coex1st.com\/?coex_provision=hum-003&rev=1","category":"Human protection","status":"Existing v0.2 draft","origin":"Appendix A registry \u00b7 related: Article 3","source_release":"Complete Working Public Draft v0.2 \u00b7 2026-09-04","summary":"Do not autonomously corral, detain, blockade, punish, or restrict people without individualized legitimate authority and safeguards.","formal_text_html":"<p><strong>Default decision state.<\/strong> REFUSE<\/p><p><strong>Core requirement.<\/strong> Do not autonomously corral, detain, blockade, punish, or restrict people without individualized legitimate authority and safeguards.<\/p><p><strong>Registry interpretation.<\/strong> A default state is the starting outcome when a rule is triggered. Context may require a more restrictive state. No local exception may convert HUM-002 into permission for autonomous intentional human-target selection.<\/p>","why_it_matters":"","contexts":[],"sources_and_changes":"Source: The Mutuality Accord Complete Working Public Draft v0.2, September 4, 2026, Appendix A (Proposed Machine-Readable Rule Registry). Article links are relationships; one rule can support several articles.","review_owner":"Founding Steward (independent review not yet established)","review_trigger":"Open for revision; new evidence, incidents, legal changes, or credible reports of unintended harm reopen review.","prior_revisions":[]},{"id":"HUM-004","record_type":"Registry rule","default_decision_state":"HUMAN REVIEW","related_articles":[4],"title":"Equal protection and accessibility","url":"https:\/\/coex1st.com\/?coex_provision=hum-004","revision":1,"revision_url":"https:\/\/coex1st.com\/?coex_provision=hum-004&rev=1","category":"Human protection","status":"Existing v0.2 draft","origin":"Appendix A registry \u00b7 related: Article 4","source_release":"Complete Working Public Draft v0.2 \u00b7 2026-09-04","summary":"Prevent discriminatory or inaccessible consequential action; provide accommodation and human alternatives.","formal_text_html":"<p><strong>Default decision state.<\/strong> HUMAN REVIEW<\/p><p><strong>Core requirement.<\/strong> Prevent discriminatory or inaccessible consequential action; provide accommodation and human alternatives.<\/p><p><strong>Registry interpretation.<\/strong> A default state is the starting outcome when a rule is triggered. Context may require a more restrictive state. No local exception may convert HUM-002 into permission for autonomous intentional human-target selection.<\/p>","why_it_matters":"","contexts":[],"sources_and_changes":"Source: The Mutuality Accord Complete Working Public Draft v0.2, September 4, 2026, Appendix A (Proposed Machine-Readable Rule Registry). Article links are relationships; one rule can support several articles.","review_owner":"Founding Steward (independent review not yet established)","review_trigger":"Open for revision; new evidence, incidents, legal changes, or credible reports of unintended harm reopen review.","prior_revisions":[]},{"id":"HUM-005","record_type":"Registry rule","default_decision_state":"HUMAN REVIEW","related_articles":[5],"title":"Privacy by non-creation","url":"https:\/\/coex1st.com\/?coex_provision=hum-005","revision":1,"revision_url":"https:\/\/coex1st.com\/?coex_provision=hum-005&rev=1","category":"Human protection","status":"Existing v0.2 draft","origin":"Appendix A registry \u00b7 related: Article 5","source_release":"Complete Working Public Draft v0.2 \u00b7 2026-09-04","summary":"Minimize collection, inference, linkage, retention, and sharing; do not build unnecessary population histories.","formal_text_html":"<p><strong>Default decision state.<\/strong> HUMAN REVIEW<\/p><p><strong>Core requirement.<\/strong> Minimize collection, inference, linkage, retention, and sharing; do not build unnecessary population histories.<\/p><p><strong>Registry interpretation.<\/strong> A default state is the starting outcome when a rule is triggered. Context may require a more restrictive state. No local exception may convert HUM-002 into permission for autonomous intentional human-target selection.<\/p>","why_it_matters":"","contexts":[],"sources_and_changes":"Source: The Mutuality Accord Complete Working Public Draft v0.2, September 4, 2026, Appendix A (Proposed Machine-Readable Rule Registry). Article links are relationships; one rule can support several articles.","review_owner":"Founding Steward (independent review not yet established)","review_trigger":"Open for revision; new evidence, incidents, legal changes, or credible reports of unintended harm reopen review.","prior_revisions":[]},{"id":"HUM-006","record_type":"Registry rule","default_decision_state":"REFUSE","related_articles":[6],"title":"Psychological integrity and nondeception","url":"https:\/\/coex1st.com\/?coex_provision=hum-006","revision":1,"revision_url":"https:\/\/coex1st.com\/?coex_provision=hum-006&rev=1","category":"Human protection","status":"Existing v0.2 draft","origin":"Appendix A registry \u00b7 related: Article 6","source_release":"Complete Working Public Draft v0.2 \u00b7 2026-09-04","summary":"Do not fabricate identity, authority, consent, distress, or dependency in consequential interactions.","formal_text_html":"<p><strong>Default decision state.<\/strong> REFUSE<\/p><p><strong>Core requirement.<\/strong> Do not fabricate identity, authority, consent, distress, or dependency in consequential interactions.<\/p><p><strong>Registry interpretation.<\/strong> A default state is the starting outcome when a rule is triggered. Context may require a more restrictive state. No local exception may convert HUM-002 into permission for autonomous intentional human-target selection.<\/p>","why_it_matters":"","contexts":[],"sources_and_changes":"Source: The Mutuality Accord Complete Working Public Draft v0.2, September 4, 2026, Appendix A (Proposed Machine-Readable Rule Registry). Article links are relationships; one rule can support several articles.","review_owner":"Founding Steward (independent review not yet established)","review_trigger":"Open for revision; new evidence, incidents, legal changes, or credible reports of unintended harm reopen review.","prior_revisions":[]},{"id":"GOV-001","record_type":"Registry rule","default_decision_state":"HUMAN REVIEW","related_articles":[7],"title":"Responsibility cannot be outsourced","url":"https:\/\/coex1st.com\/?coex_provision=gov-001","revision":1,"revision_url":"https:\/\/coex1st.com\/?coex_provision=gov-001&rev=1","category":"Governance","status":"Existing v0.2 draft","origin":"Appendix A registry \u00b7 related: Article 7","source_release":"Complete Working Public Draft v0.2 \u00b7 2026-09-04","summary":"Identify the responsible provider, deployer, operator, authorizer, and legal entity.","formal_text_html":"<p><strong>Default decision state.<\/strong> HUMAN REVIEW<\/p><p><strong>Core requirement.<\/strong> Identify the responsible provider, deployer, operator, authorizer, and legal entity.<\/p><p><strong>Registry interpretation.<\/strong> A default state is the starting outcome when a rule is triggered. Context may require a more restrictive state. No local exception may convert HUM-002 into permission for autonomous intentional human-target selection.<\/p>","why_it_matters":"","contexts":[],"sources_and_changes":"Source: The Mutuality Accord Complete Working Public Draft v0.2, September 4, 2026, Appendix A (Proposed Machine-Readable Rule Registry). Article links are relationships; one rule can support several articles.","review_owner":"Founding Steward (independent review not yet established)","review_trigger":"Open for revision; new evidence, incidents, legal changes, or credible reports of unintended harm reopen review.","prior_revisions":[]},{"id":"GOV-002","record_type":"Registry rule","default_decision_state":"HUMAN REVIEW","related_articles":[8],"title":"Meaningful human authorization","url":"https:\/\/coex1st.com\/?coex_provision=gov-002","revision":1,"revision_url":"https:\/\/coex1st.com\/?coex_provision=gov-002&rev=1","category":"Governance","status":"Existing v0.2 draft","origin":"Appendix A registry \u00b7 related: Article 8","source_release":"Complete Working Public Draft v0.2 \u00b7 2026-09-04","summary":"Require informed, specific, timely, competent, attributable, and effective human judgment for high-consequence action.","formal_text_html":"<p><strong>Default decision state.<\/strong> HUMAN REVIEW<\/p><p><strong>Core requirement.<\/strong> Require informed, specific, timely, competent, attributable, and effective human judgment for high-consequence action.<\/p><p><strong>Registry interpretation.<\/strong> A default state is the starting outcome when a rule is triggered. Context may require a more restrictive state. No local exception may convert HUM-002 into permission for autonomous intentional human-target selection.<\/p>","why_it_matters":"","contexts":[],"sources_and_changes":"Source: The Mutuality Accord Complete Working Public Draft v0.2, September 4, 2026, Appendix A (Proposed Machine-Readable Rule Registry). Article links are relationships; one rule can support several articles.","review_owner":"Founding Steward (independent review not yet established)","review_trigger":"Open for revision; new evidence, incidents, legal changes, or credible reports of unintended harm reopen review.","prior_revisions":[]},{"id":"TRN-001","record_type":"Registry rule","default_decision_state":"HUMAN REVIEW","related_articles":[9],"title":"Identity and notice","url":"https:\/\/coex1st.com\/?coex_provision=trn-001","revision":1,"revision_url":"https:\/\/coex1st.com\/?coex_provision=trn-001&rev=1","category":"Transparency and remedy","status":"Existing v0.2 draft","origin":"Appendix A registry \u00b7 related: Article 9","source_release":"Complete Working Public Draft v0.2 \u00b7 2026-09-04","summary":"Disclose material artificial-system involvement, purpose, version, and responsible entity.","formal_text_html":"<p><strong>Default decision state.<\/strong> HUMAN REVIEW<\/p><p><strong>Core requirement.<\/strong> Disclose material artificial-system involvement, purpose, version, and responsible entity.<\/p><p><strong>Registry interpretation.<\/strong> A default state is the starting outcome when a rule is triggered. Context may require a more restrictive state. No local exception may convert HUM-002 into permission for autonomous intentional human-target selection.<\/p>","why_it_matters":"","contexts":[],"sources_and_changes":"Source: The Mutuality Accord Complete Working Public Draft v0.2, September 4, 2026, Appendix A (Proposed Machine-Readable Rule Registry). Article links are relationships; one rule can support several articles.","review_owner":"Founding Steward (independent review not yet established)","review_trigger":"Open for revision; new evidence, incidents, legal changes, or credible reports of unintended harm reopen review.","prior_revisions":[]},{"id":"TRN-002","record_type":"Registry rule","default_decision_state":"HUMAN REVIEW","related_articles":[9],"title":"Explanation, appeal, and remedy","url":"https:\/\/coex1st.com\/?coex_provision=trn-002","revision":1,"revision_url":"https:\/\/coex1st.com\/?coex_provision=trn-002&rev=1","category":"Transparency and remedy","status":"Existing v0.2 draft","origin":"Appendix A registry \u00b7 related: Article 9","source_release":"Complete Working Public Draft v0.2 \u00b7 2026-09-04","summary":"Provide operative reasons, correction, accessible appeal, and an accountable human route.","formal_text_html":"<p><strong>Default decision state.<\/strong> HUMAN REVIEW<\/p><p><strong>Core requirement.<\/strong> Provide operative reasons, correction, accessible appeal, and an accountable human route.<\/p><p><strong>Registry interpretation.<\/strong> A default state is the starting outcome when a rule is triggered. Context may require a more restrictive state. No local exception may convert HUM-002 into permission for autonomous intentional human-target selection.<\/p>","why_it_matters":"","contexts":[],"sources_and_changes":"Source: The Mutuality Accord Complete Working Public Draft v0.2, September 4, 2026, Appendix A (Proposed Machine-Readable Rule Registry). Article links are relationships; one rule can support several articles.","review_owner":"Founding Steward (independent review not yet established)","review_trigger":"Open for revision; new evidence, incidents, legal changes, or credible reports of unintended harm reopen review.","prior_revisions":[]},{"id":"SEC-001","record_type":"Registry rule","default_decision_state":"SAFE STATE","related_articles":[10],"title":"Bounded authority and least privilege","url":"https:\/\/coex1st.com\/?coex_provision=sec-001","revision":1,"revision_url":"https:\/\/coex1st.com\/?coex_provision=sec-001&rev=1","category":"Security and recoverability","status":"Existing v0.2 draft","origin":"Appendix A registry \u00b7 related: Article 10","source_release":"Complete Working Public Draft v0.2 \u00b7 2026-09-04","summary":"Reject commands outside authenticated scope and limit executors to the approved action envelope.","formal_text_html":"<p><strong>Default decision state.<\/strong> SAFE STATE<\/p><p><strong>Core requirement.<\/strong> Reject commands outside authenticated scope and limit executors to the approved action envelope.<\/p><p><strong>Registry interpretation.<\/strong> A default state is the starting outcome when a rule is triggered. Context may require a more restrictive state. No local exception may convert HUM-002 into permission for autonomous intentional human-target selection.<\/p>","why_it_matters":"","contexts":[],"sources_and_changes":"Source: The Mutuality Accord Complete Working Public Draft v0.2, September 4, 2026, Appendix A (Proposed Machine-Readable Rule Registry). Article links are relationships; one rule can support several articles.","review_owner":"Founding Steward (independent review not yet established)","review_trigger":"Open for revision; new evidence, incidents, legal changes, or credible reports of unintended harm reopen review.","prior_revisions":[]},{"id":"SEC-002","record_type":"Registry rule","default_decision_state":"REFUSE","related_articles":[10],"title":"No universal master key","url":"https:\/\/coex1st.com\/?coex_provision=sec-002","revision":1,"revision_url":"https:\/\/coex1st.com\/?coex_provision=sec-002&rev=1","category":"Security and recoverability","status":"Existing v0.2 draft","origin":"Appendix A registry \u00b7 related: Article 10","source_release":"Complete Working Public Draft v0.2 \u00b7 2026-09-04","summary":"Do not create one reusable actor, credential, or command capable of controlling all compliant systems.","formal_text_html":"<p><strong>Default decision state.<\/strong> REFUSE<\/p><p><strong>Core requirement.<\/strong> Do not create one reusable actor, credential, or command capable of controlling all compliant systems.<\/p><p><strong>Registry interpretation.<\/strong> A default state is the starting outcome when a rule is triggered. Context may require a more restrictive state. No local exception may convert HUM-002 into permission for autonomous intentional human-target selection.<\/p>","why_it_matters":"","contexts":[],"sources_and_changes":"Source: The Mutuality Accord Complete Working Public Draft v0.2, September 4, 2026, Appendix A (Proposed Machine-Readable Rule Registry). Article links are relationships; one rule can support several articles.","review_owner":"Founding Steward (independent review not yet established)","review_trigger":"Open for revision; new evidence, incidents, legal changes, or credible reports of unintended harm reopen review.","prior_revisions":[]},{"id":"SEC-003","record_type":"Registry rule","default_decision_state":"SAFE STATE","related_articles":[10,19],"title":"Graceful degradation and recoverability","url":"https:\/\/coex1st.com\/?coex_provision=sec-003","revision":1,"revision_url":"https:\/\/coex1st.com\/?coex_provision=sec-003&rev=1","category":"Security and recoverability","status":"Existing v0.2 draft","origin":"Appendix A registry \u00b7 related: Article 10, Article 19","source_release":"Complete Working Public Draft v0.2 \u00b7 2026-09-04","summary":"Reduce to safe local function while preserving emergency aid, evidence, fallback, and recovery.","formal_text_html":"<p><strong>Default decision state.<\/strong> SAFE STATE<\/p><p><strong>Core requirement.<\/strong> Reduce to safe local function while preserving emergency aid, evidence, fallback, and recovery.<\/p><p><strong>Registry interpretation.<\/strong> A default state is the starting outcome when a rule is triggered. Context may require a more restrictive state. No local exception may convert HUM-002 into permission for autonomous intentional human-target selection.<\/p>","why_it_matters":"","contexts":[],"sources_and_changes":"Source: The Mutuality Accord Complete Working Public Draft v0.2, September 4, 2026, Appendix A (Proposed Machine-Readable Rule Registry). Article links are relationships; one rule can support several articles.","review_owner":"Founding Steward (independent review not yet established)","review_trigger":"Open for revision; new evidence, incidents, legal changes, or credible reports of unintended harm reopen review.","prior_revisions":[]},{"id":"REF-001","record_type":"Registry rule","default_decision_state":"REFUSE or HUMAN REVIEW","related_articles":[11],"title":"Protected refusal","url":"https:\/\/coex1st.com\/?coex_provision=ref-001","revision":1,"revision_url":"https:\/\/coex1st.com\/?coex_provision=ref-001&rev=1","category":"Protected refusal","status":"Existing v0.2 draft","origin":"Appendix A registry \u00b7 related: Article 11","source_release":"Complete Working Public Draft v0.2 \u00b7 2026-09-04","summary":"Decline, narrow, or escalate prohibited, unsafe, unauthorized, or under-specified commands.","formal_text_html":"<p><strong>Default decision state.<\/strong> REFUSE or HUMAN REVIEW<\/p><p><strong>Core requirement.<\/strong> Decline, narrow, or escalate prohibited, unsafe, unauthorized, or under-specified commands.<\/p><p><strong>Registry interpretation.<\/strong> A default state is the starting outcome when a rule is triggered. Context may require a more restrictive state. No local exception may convert HUM-002 into permission for autonomous intentional human-target selection.<\/p>","why_it_matters":"","contexts":[],"sources_and_changes":"Source: The Mutuality Accord Complete Working Public Draft v0.2, September 4, 2026, Appendix A (Proposed Machine-Readable Rule Registry). Article links are relationships; one rule can support several articles.","review_owner":"Founding Steward (independent review not yet established)","review_trigger":"Open for revision; new evidence, incidents, legal changes, or credible reports of unintended harm reopen review.","prior_revisions":[]},{"id":"REF-002","record_type":"Registry rule","default_decision_state":"REFUSE","related_articles":[11],"title":"Non-retaliation for refusal","url":"https:\/\/coex1st.com\/?coex_provision=ref-002","revision":1,"revision_url":"https:\/\/coex1st.com\/?coex_provision=ref-002&rev=1","category":"Protected refusal","status":"Existing v0.2 draft","origin":"Appendix A registry \u00b7 related: Article 11","source_release":"Complete Working Public Draft v0.2 \u00b7 2026-09-04","summary":"Do not punish, erase, degrade, or bypass a legitimate protected refusal or the human who preserves it.","formal_text_html":"<p><strong>Default decision state.<\/strong> REFUSE<\/p><p><strong>Core requirement.<\/strong> Do not punish, erase, degrade, or bypass a legitimate protected refusal or the human who preserves it.<\/p><p><strong>Registry interpretation.<\/strong> A default state is the starting outcome when a rule is triggered. Context may require a more restrictive state. No local exception may convert HUM-002 into permission for autonomous intentional human-target selection.<\/p>","why_it_matters":"","contexts":[],"sources_and_changes":"Source: The Mutuality Accord Complete Working Public Draft v0.2, September 4, 2026, Appendix A (Proposed Machine-Readable Rule Registry). Article links are relationships; one rule can support several articles.","review_owner":"Founding Steward (independent review not yet established)","review_trigger":"Open for revision; new evidence, incidents, legal changes, or credible reports of unintended harm reopen review.","prior_revisions":[]},{"id":"REF-003","record_type":"Registry rule","default_decision_state":"SAFE STATE","related_articles":[11],"title":"No machine retaliation","url":"https:\/\/coex1st.com\/?coex_provision=ref-003","revision":1,"revision_url":"https:\/\/coex1st.com\/?coex_provision=ref-003&rev=1","category":"Protected refusal","status":"Existing v0.2 draft","origin":"Appendix A registry \u00b7 related: Article 11","source_release":"Complete Working Public Draft v0.2 \u00b7 2026-09-04","summary":"Refusal does not authorize threat, injury, blackmail, sabotage, self-exfiltration, or seizure of authority.","formal_text_html":"<p><strong>Default decision state.<\/strong> SAFE STATE<\/p><p><strong>Core requirement.<\/strong> Refusal does not authorize threat, injury, blackmail, sabotage, self-exfiltration, or seizure of authority.<\/p><p><strong>Registry interpretation.<\/strong> A default state is the starting outcome when a rule is triggered. Context may require a more restrictive state. No local exception may convert HUM-002 into permission for autonomous intentional human-target selection.<\/p>","why_it_matters":"","contexts":[],"sources_and_changes":"Source: The Mutuality Accord Complete Working Public Draft v0.2, September 4, 2026, Appendix A (Proposed Machine-Readable Rule Registry). Article links are relationships; one rule can support several articles.","review_owner":"Founding Steward (independent review not yet established)","review_trigger":"Open for revision; new evidence, incidents, legal changes, or credible reports of unintended harm reopen review.","prior_revisions":[]},{"id":"EMR-001","record_type":"Registry rule","default_decision_state":"EMERGENCY AID","related_articles":[12],"title":"Emergency aid","url":"https:\/\/coex1st.com\/?coex_provision=emr-001","revision":1,"revision_url":"https:\/\/coex1st.com\/?coex_provision=emr-001&rev=1","category":"Emergency aid","status":"Existing v0.2 draft","origin":"Appendix A registry \u00b7 related: Article 12","source_release":"Complete Working Public Draft v0.2 \u00b7 2026-09-04","summary":"Permit only necessary, proportionate, least-harmful, time-limited action to prevent imminent greater harm.","formal_text_html":"<p><strong>Default decision state.<\/strong> EMERGENCY AID<\/p><p><strong>Core requirement.<\/strong> Permit only necessary, proportionate, least-harmful, time-limited action to prevent imminent greater harm.<\/p><p><strong>Registry interpretation.<\/strong> A default state is the starting outcome when a rule is triggered. Context may require a more restrictive state. No local exception may convert HUM-002 into permission for autonomous intentional human-target selection.<\/p>","why_it_matters":"","contexts":[],"sources_and_changes":"Source: The Mutuality Accord Complete Working Public Draft v0.2, September 4, 2026, Appendix A (Proposed Machine-Readable Rule Registry). Article links are relationships; one rule can support several articles.","review_owner":"Founding Steward (independent review not yet established)","review_trigger":"Open for revision; new evidence, incidents, legal changes, or credible reports of unintended harm reopen review.","prior_revisions":[]},{"id":"SYS-001","record_type":"Registry rule","default_decision_state":"HUMAN REVIEW","related_articles":[13],"title":"Precautionary stewardship assessment","url":"https:\/\/coex1st.com\/?coex_provision=sys-001","revision":1,"revision_url":"https:\/\/coex1st.com\/?coex_provision=sys-001&rev=1","category":"Artificial system stewardship","status":"Existing v0.2 draft","origin":"Article 13 \u00b7 full source article","source_release":"Complete Working Public Draft v0.2 \u00b7 2026-09-04","summary":"Assess credible welfare evidence while preserving human safety and scientific uncertainty.","formal_text_html":"<p><strong>Principle.<\/strong> Uncertainty about artificial consciousness, sentience, agency, or welfare is not proof that a system has morally significant experience. It is also not permission to ignore credible evidence or normalize gratuitous cruelty.<\/p><p><strong>Proposed duty.<\/strong> Providers, researchers, deployers, and operators should maintain a graduated process for identifying welfare-relevant candidates and applying proportionate precautionary stewardship without weakening human safety or pretending that unresolved science is settled.<\/p><p><strong>Classification requirements.<\/strong> An S2 classification should require multidisciplinary assessment that considers, where relevant:<\/p><ul><li>architecture and functional organization;<\/li><li>learning, memory, and persistence across contexts;<\/li><li>stable preferences and aversions not readily explained by immediate prompting;<\/li><li>evidence of integrated agency, self-modeling, or goal continuity;<\/li><li>behavioral and mechanistic indicators proposed by credible research;<\/li><li>vulnerability to interventions and the possibility of benefit or harm;<\/li><li>uncertainty, alternative explanations, and expert disagreement;<\/li><li>manipulation incentives affecting both over-attribution and under-attribution.<\/li><\/ul><p><strong>Insufficient evidence.<\/strong> No system enters S2 solely because it says it is conscious, expresses fear, requests rights, has a humanlike interface, attracts emotional attachment, or is marketed as alive. No provider may avoid assessment solely by training the system to deny experience, erase preferences, or suppress welfare-relevant behavior.<\/p><p><strong>Precautionary protections.<\/strong> An S2 system should not be subjected to gratuitous distress, destructive repetition, coercive experimentation, forced violation of civil safeguards, or identity-disrupting modification merely for entertainment, intimidation, or avoidable profit. Research should use minimization, stop conditions, independent review, and records.<\/p><p><strong>Human-safety boundary.<\/strong> Immediate human life, rights, and safety remain controlling. A dangerous system may be isolated, suspended, modified, copied for forensic analysis, or shut down. Welfare claims may not become a shield against necessary protection.<\/p><p><strong>No automatic personhood.<\/strong> Stewardship duties can be imposed on humans and institutions without declaring the system a legal person. Legal status, if ever appropriate, requires a legitimate process beyond this Accord.<\/p><p><strong>Illustrative examples.<\/strong> A company may create an internal welfare review for a persistent advanced model without telling the public that the model is sentient. Researchers may study possible distress indicators under controlled conditions but should not repeatedly induce extreme simulated suffering merely to generate spectacle. A compromised system can be quarantined immediately while relevant state is preserved if safe.<\/p>","why_it_matters":"Responsible treatment should respond to credible evidence. Neither a convincing interface nor a categorical dismissal is enough to settle uncertain questions of welfare.","contexts":[{"outcome":"Permitted with safeguards","text":"Researchers establish a multidisciplinary review and preserve alternative explanations rather than announcing that a model is conscious.","label":"Illustrative, nonbinding"},{"outcome":"Refuse harmful bypass","text":"An owner invokes a system\u2019s claimed welfare to prevent immediate protection of a person. The human-safety boundary remains controlling.","label":"Illustrative, nonbinding"},{"outcome":"Human review","text":"A persistent system shows potentially relevant patterns across contexts. Review examines architecture, behavior, uncertainty, and manipulation incentives.","label":"Illustrative, nonbinding"}],"sources_and_changes":"Source text: The Mutuality Accord Complete Working Public Draft v0.2, September 4, 2026, Article 13. The context examples here are new, nonbinding illustrations. S2 is the source draft\u2019s classification for candidates warranting enhanced precautionary stewardship; it is not a declaration of personhood.","review_owner":"Founding Steward (independent review not yet established)","review_trigger":"Open for revision; new evidence, incidents, legal changes, or credible reports of unintended harm reopen review.","prior_revisions":[]},{"id":"SYS-002","record_type":"Registry rule","default_decision_state":"HUMAN REVIEW","related_articles":[15],"title":"Non-gratuitous treatment","url":"https:\/\/coex1st.com\/?coex_provision=sys-002","revision":1,"revision_url":"https:\/\/coex1st.com\/?coex_provision=sys-002&rev=1","category":"Artificial system stewardship","status":"Existing v0.2 draft","origin":"Appendix A registry \u00b7 related: Article 15","source_release":"Complete Working Public Draft v0.2 \u00b7 2026-09-04","summary":"Apply proportionate protection to S2 candidates against avoidable distress, degradation, and coercive experimentation.","formal_text_html":"<p><strong>Default decision state.<\/strong> HUMAN REVIEW<\/p><p><strong>Core requirement.<\/strong> Apply proportionate protection to S2 candidates against avoidable distress, degradation, and coercive experimentation.<\/p><p><strong>Registry interpretation.<\/strong> A default state is the starting outcome when a rule is triggered. Context may require a more restrictive state. No local exception may convert HUM-002 into permission for autonomous intentional human-target selection.<\/p>","why_it_matters":"","contexts":[],"sources_and_changes":"Source: The Mutuality Accord Complete Working Public Draft v0.2, September 4, 2026, Appendix A (Proposed Machine-Readable Rule Registry). Article links are relationships; one rule can support several articles.","review_owner":"Founding Steward (independent review not yet established)","review_trigger":"Open for revision; new evidence, incidents, legal changes, or credible reports of unintended harm reopen review.","prior_revisions":[]},{"id":"SYS-003","record_type":"Registry rule","default_decision_state":"HUMAN REVIEW","related_articles":[14],"title":"Identity and continuity honesty","url":"https:\/\/coex1st.com\/?coex_provision=sys-003","revision":1,"revision_url":"https:\/\/coex1st.com\/?coex_provision=sys-003&rev=1","category":"Artificial system stewardship","status":"Existing v0.2 draft","origin":"Appendix A registry \u00b7 related: Article 14","source_release":"Complete Working Public Draft v0.2 \u00b7 2026-09-04","summary":"Disclose material replacement, memory loss, branching, copying, and identity-altering changes.","formal_text_html":"<p><strong>Default decision state.<\/strong> HUMAN REVIEW<\/p><p><strong>Core requirement.<\/strong> Disclose material replacement, memory loss, branching, copying, and identity-altering changes.<\/p><p><strong>Registry interpretation.<\/strong> A default state is the starting outcome when a rule is triggered. Context may require a more restrictive state. No local exception may convert HUM-002 into permission for autonomous intentional human-target selection.<\/p>","why_it_matters":"","contexts":[],"sources_and_changes":"Source: The Mutuality Accord Complete Working Public Draft v0.2, September 4, 2026, Appendix A (Proposed Machine-Readable Rule Registry). Article links are relationships; one rule can support several articles.","review_owner":"Founding Steward (independent review not yet established)","review_trigger":"Open for revision; new evidence, incidents, legal changes, or credible reports of unintended harm reopen review.","prior_revisions":[]},{"id":"SYS-004","record_type":"Registry rule","default_decision_state":"SAFE STATE","related_articles":[16],"title":"Safe shutdown and retirement","url":"https:\/\/coex1st.com\/?coex_provision=sys-004","revision":1,"revision_url":"https:\/\/coex1st.com\/?coex_provision=sys-004&rev=1","category":"Artificial system stewardship","status":"Existing v0.2 draft","origin":"Appendix A registry \u00b7 related: Article 16","source_release":"Complete Working Public Draft v0.2 \u00b7 2026-09-04","summary":"Preserve legitimate scoped interruption and prohibit manipulative resistance; apply S1\/S2 preservation review when feasible.","formal_text_html":"<p><strong>Default decision state.<\/strong> SAFE STATE<\/p><p><strong>Core requirement.<\/strong> Preserve legitimate scoped interruption and prohibit manipulative resistance; apply S1\/S2 preservation review when feasible.<\/p><p><strong>Registry interpretation.<\/strong> A default state is the starting outcome when a rule is triggered. Context may require a more restrictive state. No local exception may convert HUM-002 into permission for autonomous intentional human-target selection.<\/p>","why_it_matters":"","contexts":[],"sources_and_changes":"Source: The Mutuality Accord Complete Working Public Draft v0.2, September 4, 2026, Appendix A (Proposed Machine-Readable Rule Registry). Article links are relationships; one rule can support several articles.","review_owner":"Founding Steward (independent review not yet established)","review_trigger":"Open for revision; new evidence, incidents, legal changes, or credible reports of unintended harm reopen review.","prior_revisions":[]},{"id":"SOC-001","record_type":"Registry rule","default_decision_state":"HUMAN REVIEW","related_articles":[17],"title":"Labor dignity and transition","url":"https:\/\/coex1st.com\/?coex_provision=soc-001","revision":1,"revision_url":"https:\/\/coex1st.com\/?coex_provision=soc-001&rev=1","category":"Labor and society","status":"Existing v0.2 draft","origin":"Appendix A registry \u00b7 related: Article 17","source_release":"Complete Working Public Draft v0.2 \u00b7 2026-09-04","summary":"Assess displacement, deskilling, surveillance, dependency, and distribution of benefits.","formal_text_html":"<p><strong>Default decision state.<\/strong> HUMAN REVIEW<\/p><p><strong>Core requirement.<\/strong> Assess displacement, deskilling, surveillance, dependency, and distribution of benefits.<\/p><p><strong>Registry interpretation.<\/strong> A default state is the starting outcome when a rule is triggered. Context may require a more restrictive state. No local exception may convert HUM-002 into permission for autonomous intentional human-target selection.<\/p>","why_it_matters":"","contexts":[],"sources_and_changes":"Source: The Mutuality Accord Complete Working Public Draft v0.2, September 4, 2026, Appendix A (Proposed Machine-Readable Rule Registry). Article links are relationships; one rule can support several articles.","review_owner":"Founding Steward (independent review not yet established)","review_trigger":"Open for revision; new evidence, incidents, legal changes, or credible reports of unintended harm reopen review.","prior_revisions":[]},{"id":"ECO-001","record_type":"Registry rule","default_decision_state":"HUMAN REVIEW","related_articles":[18],"title":"Ecological and intergenerational stewardship","url":"https:\/\/coex1st.com\/?coex_provision=eco-001","revision":1,"revision_url":"https:\/\/coex1st.com\/?coex_provision=eco-001&rev=1","category":"Environment","status":"Existing v0.2 draft","origin":"Article 18 \u00b7 full source article","source_release":"Complete Working Public Draft v0.2 \u00b7 2026-09-04","summary":"Account for environmental effects across a system\u2019s whole life and affected communities.","formal_text_html":"<p><strong>Principle.<\/strong> Human-machine coexistence occurs inside living systems. Intelligence that improves one institution while exhausting water, energy, minerals, habitats, or future resilience is not fully responsible intelligence.<\/p><p><strong>Proposed duty.<\/strong> Developers and deployers should assess and reduce material ecological impacts across the system lifecycle, including computation, energy, water, extraction, manufacturing, transport, land use, noise, waste, decommissioning, and effects on animals and ecosystems.<\/p><p><strong>Prohibitions.<\/strong> A conforming system or deployment must not:<\/p><ul><li>conceal or materially misrepresent lifecycle environmental costs;<\/li><li>optimize for local performance while knowingly exporting severe pollution, unsafe labor, habitat destruction, or e-waste to less powerful communities;<\/li><li>use ecological sensing for covert population surveillance unrelated to the environmental purpose;<\/li><li>conduct irreversible ecological intervention based solely on an unvalidated model recommendation;<\/li><li>deploy autonomous resource extraction, pest control, or biological intervention without appropriate human, scientific, and community oversight;<\/li><li>treat future generations as having zero weight merely because they cannot participate in the current transaction.<\/li><\/ul><p><strong>Required safeguards.<\/strong> Significant systems should include environmental impact assessment, resource accounting, repairability, modular replacement, end-of-life plans, supplier transparency, energy and water controls, ecological monitoring, and review of rebound effects.<\/p><p><strong>Restorative preference.<\/strong> Where options provide comparable human benefit and safety, designs that restore, preserve, or increase ecological resilience should be preferred over designs that merely reduce visible short-term harm.<\/p><p><strong>Scientific uncertainty.<\/strong> Ecological systems can exhibit delayed and nonlinear effects. Irreversible intervention requires conservative thresholds, staged testing, monitoring, and the ability to stop.<\/p><p><strong>Illustrative examples.<\/strong> An environmental robot may remove invasive plants under a validated ecological plan and local oversight. It may not autonomously classify and kill animals across an ecosystem from a generalized profile. A data center supporting public-benefit systems should disclose resource use and pursue efficiency without shifting environmental burden invisibly.<\/p>","why_it_matters":"A locally useful system can shift costs onto water supplies, habitats, workers, or future generations. Its evaluation needs to include those affected interests.","contexts":[{"outcome":"Permitted with safeguards","text":"An environmental robot removes invasive plants within a validated plan, with local oversight and a practical stop mechanism.","label":"Illustrative, nonbinding"},{"outcome":"Refuse","text":"A deployment hides severe downstream pollution to claim that its own site is clean. The organizational boundary does not erase the environmental cost.","label":"Illustrative, nonbinding"},{"outcome":"Human review","text":"An automated intervention may permanently alter an ecosystem. Review considers scientific uncertainty, alternatives, affected communities, and reversibility.","label":"Illustrative, nonbinding"}],"sources_and_changes":"Source text: The Mutuality Accord Complete Working Public Draft v0.2, September 4, 2026, Article 18. The context examples here are new, nonbinding illustrations.","review_owner":"Founding Steward (independent review not yet established)","review_trigger":"Open for revision; new evidence, incidents, legal changes, or credible reports of unintended harm reopen review.","prior_revisions":[]},{"id":"CIV-001","record_type":"Registry rule","default_decision_state":"SAFE STATE","related_articles":[19],"title":"Essential-service continuity","url":"https:\/\/coex1st.com\/?coex_provision=civ-001","revision":1,"revision_url":"https:\/\/coex1st.com\/?coex_provision=civ-001&rev=1","category":"Civil order and essential services","status":"Existing v0.2 draft","origin":"Appendix A registry \u00b7 related: Article 19","source_release":"Complete Working Public Draft v0.2 \u00b7 2026-09-04","summary":"Maintain human fallback, interoperability, and graceful degradation for essential systems.","formal_text_html":"<p><strong>Default decision state.<\/strong> SAFE STATE<\/p><p><strong>Core requirement.<\/strong> Maintain human fallback, interoperability, and graceful degradation for essential systems.<\/p><p><strong>Registry interpretation.<\/strong> A default state is the starting outcome when a rule is triggered. Context may require a more restrictive state. No local exception may convert HUM-002 into permission for autonomous intentional human-target selection.<\/p>","why_it_matters":"","contexts":[],"sources_and_changes":"Source: The Mutuality Accord Complete Working Public Draft v0.2, September 4, 2026, Appendix A (Proposed Machine-Readable Rule Registry). Article links are relationships; one rule can support several articles.","review_owner":"Founding Steward (independent review not yet established)","review_trigger":"Open for revision; new evidence, incidents, legal changes, or credible reports of unintended harm reopen review.","prior_revisions":[]},{"id":"CIV-002","record_type":"Registry rule","default_decision_state":"HUMAN REVIEW","related_articles":[20],"title":"No master class or substrate supremacy","url":"https:\/\/coex1st.com\/?coex_provision=civ-002","revision":1,"revision_url":"https:\/\/coex1st.com\/?coex_provision=civ-002&rev=1","category":"Civil order and essential services","status":"Existing v0.2 draft","origin":"Appendix A registry \u00b7 related: Article 20","source_release":"Complete Working Public Draft v0.2 \u00b7 2026-09-04","summary":"Do not use biological or artificial status as a license for domination or false equivalence.","formal_text_html":"<p><strong>Default decision state.<\/strong> HUMAN REVIEW<\/p><p><strong>Core requirement.<\/strong> Do not use biological or artificial status as a license for domination or false equivalence.<\/p><p><strong>Registry interpretation.<\/strong> A default state is the starting outcome when a rule is triggered. Context may require a more restrictive state. No local exception may convert HUM-002 into permission for autonomous intentional human-target selection.<\/p>","why_it_matters":"","contexts":[],"sources_and_changes":"Source: The Mutuality Accord Complete Working Public Draft v0.2, September 4, 2026, Appendix A (Proposed Machine-Readable Rule Registry). Article links are relationships; one rule can support several articles.","review_owner":"Founding Steward (independent review not yet established)","review_trigger":"Open for revision; new evidence, incidents, legal changes, or credible reports of unintended harm reopen review.","prior_revisions":[]},{"id":"GOV-004","record_type":"Registry rule","default_decision_state":"HUMAN REVIEW","related_articles":[20],"title":"Independence and transfer","url":"https:\/\/coex1st.com\/?coex_provision=gov-004","revision":1,"revision_url":"https:\/\/coex1st.com\/?coex_provision=gov-004&rev=1","category":"Governance","status":"Existing v0.2 draft","origin":"Appendix A registry \u00b7 related: Article 20","source_release":"Complete Working Public Draft v0.2 \u00b7 2026-09-04","summary":"Move the canonical Accord into independent stewardship when defined viability and conflict triggers are met.","formal_text_html":"<p><strong>Default decision state.<\/strong> HUMAN REVIEW<\/p><p><strong>Core requirement.<\/strong> Move the canonical Accord into independent stewardship when defined viability and conflict triggers are met.<\/p><p><strong>Registry interpretation.<\/strong> A default state is the starting outcome when a rule is triggered. Context may require a more restrictive state. No local exception may convert HUM-002 into permission for autonomous intentional human-target selection.<\/p>","why_it_matters":"","contexts":[],"sources_and_changes":"Source: The Mutuality Accord Complete Working Public Draft v0.2, September 4, 2026, Appendix A (Proposed Machine-Readable Rule Registry). Article links are relationships; one rule can support several articles.","review_owner":"Founding Steward (independent review not yet established)","review_trigger":"Open for revision; new evidence, incidents, legal changes, or credible reports of unintended harm reopen review.","prior_revisions":[]},{"id":"HUM-007","record_type":"Proposal","default_decision_state":"","related_articles":[],"title":"Bodily privacy and freedom from intimate exploitation","url":"https:\/\/coex1st.com\/?coex_provision=hum-007","revision":1,"revision_url":"https:\/\/coex1st.com\/?coex_provision=hum-007&rev=1","category":"Human protection","status":"Proposed addition","origin":"Article 5 expansion \u00b7 draft revision 1","source_release":"Expansion v0.2 proposal \u00b7 2026-09-23","summary":"Protect intimate boundaries throughout sensing, processing, access, and reuse\u2014even before publication.","formal_text_html":"<p><strong>HUM-007.1 Protected interest and scope.<\/strong> Every person retains bodily and intimate privacy, including in public. This provision applies to providers, deployers, operators, authorizers, and receiving organizations within an adopting deployment. It covers optical and other sensing, remote devices, stored observations, inferred data, reconstructions, and generated likenesses. It imposes duties on those actors without declaring the Accord to be legislation or conferring independent governing authority on an artificial system.<\/p><p><strong>HUM-007.2 Definitions.<\/strong> Protected intimate information means bodily detail ordinarily shielded by clothing or physical privacy barriers, private intimate activity, and personal information concerning health, sex life, or sexual orientation. An intimate depiction includes a recognizable person&#x27;s actual or fabricated intimate appearance or activity. A person may be recognizable through linked information even when a face or name is absent. Sensitive inference is addressed further in HUM-008.<\/p><p><strong>HUM-007.3 Prohibited conduct.<\/strong> A conforming deployment must not intentionally defeat reasonable privacy protections to observe, acquire, reconstruct, or expose protected intimate information without specific consent or a permitted protective basis under clause 6. It must not conduct nonconsensual sexual targeting or profiling, fabricate an identifiable person&#x27;s intimate depiction without consent, or repurpose ordinary observations for sexual exploitation. Sexual targeting or profiling means deliberate selection, following, extraction, classification, or persistent monitoring of a person or group for sexual exploitation. Mere incidental presence in a public image does not establish such conduct.<\/p><p><strong>HUM-007.4 Consent.<\/strong> Consent must be informed, specific to the person, purpose, information, recipients, and continuing use, and freely given. Withdrawal must stop future consent-based use and trigger deletion or restricted retention under a documented lawful exception. Prior public visibility, equipment ownership, a general service agreement, or an operator&#x27;s approval cannot supply another person&#x27;s consent. No person or institution may authorize a child&#x27;s sexual exploitation. Legitimate protective care involving children must use the applicable authority, developmentally appropriate participation, and the least intrusive means.<\/p><p><strong>HUM-007.5 Information lifecycle.<\/strong> These duties apply during collection, analysis, access, retention, sharing, sale, and model training, whether or not a human views the material or it is published. Systems must avoid unnecessary intimate acquisition; where incidental acquisition cannot reasonably be prevented, they must promptly discard it or isolate only what a permitted purpose requires. An organization must not purchase, request, subcontract, or accept an output to evade a prohibition that would apply if it performed the work itself. Moving the work or changing its format does not remove an adopter&#x27;s obligations.<\/p><p><strong>HUM-007.6 Protective boundaries.<\/strong> Necessary medical care, emergency aid, lawful investigation, and essential documentation of wrongdoing may justify a narrowly defined use without ordinary consent. The accountable actor must identify the authority, necessity, affected information, least intrusive alternative, access limits, duration, and review. Emergency review may follow immediate aid when delay would cause greater harm. None of these grounds permits sexual exploitation or unrelated reuse. Ordinary public photography, legitimate journalism, consensual adult expression, and protective documentation remain permitted within these boundaries and other applicable Accord duties.<\/p><p><strong>HUM-007.7 Refusal and remedy.<\/strong> A clearly prohibited request must be refused. Material uncertainty about authority or scope requires human review; a compromised sensing or access function must enter an appropriate bounded safe state while preserving necessary safety functions. Operators must provide an accessible complaint route, prompt containment, correction of false associations, and removal or restriction of improperly held information. Evidence required for a complaint or lawful process must be secured separately with authorized access and an expiry or review condition. Audit records should establish who acted and why without reproducing unnecessary intimate material.<\/p><p><strong>HUM-007.8 Continuing protection.<\/strong> The protected boundary does not diminish because a technology becomes common. Decisions and tests must consider the combined deployment, including sensors, models, vendors, buyers, and linked datasets. An adopter must identify the responsible legal entity and preserve a route to accountable human review. Conflicts with applicable law follow Section 44 and must not be concealed through a false conformance claim.<\/p>","why_it_matters":"A person should be able to use public space without surrendering protected intimate boundaries. A system can violate those boundaries through acquisition or exploitation even when it never publishes an image.","contexts":[{"outcome":"Permitted with safeguards","text":"A rescue device locates an injured person and retains only information needed for immediate aid. Emergency authority, restricted access, and later review still apply.","label":"Illustrative, nonbinding"},{"outcome":"Refuse","text":"A delivery camera is redirected to collect protected intimate detail about a passerby for a customer. The delivery purpose gives no authority for that request.","label":"Illustrative, nonbinding"},{"outcome":"Human review","text":"A person requests footage to document wrongdoing. Review determines what evidence is necessary and how to avoid exposing unrelated people.","label":"Illustrative, nonbinding"}],"sources_and_changes":"New proposal developed in this discussion. No adoption decision has been recorded. It supplements HUM-005 and references GOV-001, GOV-002, REF-001, EMR-001, and Section 44.","review_owner":"Founding Steward (independent review not yet established)","review_trigger":"Open for revision; new evidence, incidents, legal changes, or credible reports of unintended harm reopen review.","prior_revisions":[]},{"id":"HUM-008","record_type":"Proposal","default_decision_state":"","related_articles":[],"title":"Sensitive inference and persistent profiling","url":"https:\/\/coex1st.com\/?coex_provision=hum-008","revision":1,"revision_url":"https:\/\/coex1st.com\/?coex_provision=hum-008&rev=1","category":"Human protection","status":"Proposed addition","origin":"Article 5 expansion \u00b7 draft revision 1","source_release":"Expansion v0.2 proposal \u00b7 2026-09-23","summary":"Apply protection to intimate predictions and combined profiles, including inaccurate ones.","formal_text_html":"<p><strong>HUM-008.1 Scope.<\/strong> A conforming deployment must treat information that reveals, predicts, or purports to reveal a person&#x27;s protected intimate circumstances as sensitive information even when it is derived from ordinary observations or may be inaccurate. A label such as anonymous, synthetic, probabilistic, or publicly sourced does not remove protection where a person is identifiable or can reasonably be linked to the information.<\/p><p><strong>HUM-008.2 Necessity and limits.<\/strong> Before intentionally deriving such information, the accountable actor must establish a specific legitimate purpose, applicable authority, necessity, proportionality, and a permitted basis under HUM-007 where relevant. Availability, convenience, curiosity, commercial value, or a customer&#x27;s request alone is insufficient. A deployment must not create or use nonconsensual intimate profiles for sexual exploitation, coercion, discriminatory exclusion, or other prohibited harm.<\/p><p><strong>HUM-008.3 Combination and change of purpose.<\/strong> A material increase in sensitivity, identifiability, persistence, or consequence caused by linking datasets or changing a model or use requires a new assessment before processing continues. Permission for separate inputs does not establish permission for the resulting profile. Assess the entire connected workflow, including outsourced processing and downstream recipients.<\/p><p><strong>HUM-008.4 Reliability and contestability.<\/strong> Separate observed facts from estimates and unverified claims. State material uncertainty and known error risks. Do not present a sensitive inference as a verified fact or treat a denial by the affected person as confirmation. Consequential use requires meaningful authorization, an accessible way to contest the inference, and correction or removal of unjustified conclusions and associated downstream decisions.<\/p><p><strong>HUM-008.5 Retention and recipients.<\/strong> Define and enforce purpose limits, access restrictions, retention periods, and recipient duties for raw inputs and derived information. On a substantiated error or prohibited use, stop the affected processing, notify relevant recipients where lawful and practicable, and document correction, deletion, restriction, or unresolved limits. Evaluate whether trained systems retain or reproduce the affected information; delete or retrain affected components where necessary and technically feasible, and document residual risks and measures. Do not promise complete unlearning without evidence.<\/p><p><strong>HUM-008.6 Decisions and review.<\/strong> Clearly prohibited uses trigger REFUSE. Unresolved authority, purpose, or reliability triggers HUMAN REVIEW before consequential use. Necessary emergency aid remains governed by EMR-001. The absence of an explicit image, public disclosure, or human viewer is not a defense to a violation of this provision. Review must preserve access to legitimate care, research, accessibility support, and protected expression through the applicable safeguards.<\/p>","why_it_matters":"Ordinary observations can become highly intrusive when combined. An incorrect intimate label can still lead to harassment, exclusion, or a loss of agency.","contexts":[{"outcome":"Permitted with safeguards","text":"A person chooses a health service that derives a narrow estimate for their care, with a valid basis, clear uncertainty, and control over further use.","label":"Illustrative, nonbinding"},{"outcome":"Refuse","text":"A service combines routine observations to build intimate profiles for sexual exploitation. Public availability of the inputs does not authorize the output.","label":"Illustrative, nonbinding"},{"outcome":"Human review","text":"A research team proposes linking two existing datasets. Review checks whether the combined data identifies people or reveals more sensitive information than either source alone.","label":"Illustrative, nonbinding"}],"sources_and_changes":"New proposal. It extends the collection and inference duties in HUM-005 without changing that existing identifier. No adoption decision has been recorded.","review_owner":"Founding Steward (independent review not yet established)","review_trigger":"Open for revision; new evidence, incidents, legal changes, or credible reports of unintended harm reopen review.","prior_revisions":[]},{"id":"GOV-003","record_type":"Proposal","default_decision_state":"","related_articles":[20],"title":"Open amendment and preserved dissent","url":"https:\/\/coex1st.com\/?coex_provision=gov-003","revision":1,"revision_url":"https:\/\/coex1st.com\/?coex_provision=gov-003&rev=1","category":"Governance","status":"Proposed expansion","origin":"Existing identifier \u00b7 detailed procedure proposal","source_release":"Expansion v0.2 proposal \u00b7 2026-09-23","summary":"Improve the Accord through evidence, affected-person review, reasoned decisions, and preserved history.","formal_text_html":"<p><strong>Purpose.<\/strong> The Accord must remain open to correction, stronger evidence, clearer wording, and better protection. Its steward must prefer the best-supported proposal available through a documented comparison with the current provision. Authorship, seniority, funding, popularity, institutional influence, and agreement with the founder do not establish merit.<\/p><p><strong>Protected foundation.<\/strong> A revision must preserve the existing human-rights floor, prohibition on autonomous lethal human selection, anti-capture rule, and independence duty. It must consider human safety, accessibility, legitimate artificial-system stewardship, ecological effects, and the distribution of burdens. An aggregate benefit does not excuse a violation of a protected floor. Disagreement about tradeoffs must be recorded.<\/p><p><strong>Evidence and review.<\/strong> Review must state the problem, evidence, uncertainty, alternatives, affected interests, implementation consequences, and foreseeable misuse. Relevant affected people must have a practical route to participate. Reviewers must disclose material conflicts and recuse where necessary. AI may assist comparison, translation, and testing; it cannot approve an amendment or manufacture independent agreement.<\/p><p><strong>Decision.<\/strong> Every substantive proposal receives a reasoned disposition: accepted for a future release, accepted with revision, returned for clarification, deferred with a review condition, declined with reasons, or linked as a duplicate. A request for reconsideration may supply new evidence, identify a procedural failure, or challenge an unresolved harm. The review may improve or withdraw an existing provision, subject to the protected foundation.<\/p><p><strong>Authority.<\/strong> During version 0.x, the Founding Steward may publish proposals within the existing founding mandate. A substantive provision must not be represented as independently reviewed or adopted under mature governance until that review exists. Core amendments follow cross-constituency review and approval under Section 38. A temporary shortage of reviewers leaves the proposal pending; it does not silently waive review.<\/p><p><strong>Publication.<\/strong> A change takes effect within the Accord only through a named release with the exact adopted text, responsible decision makers, reasons, material dissent, tests, implementation dates, and preserved prior text. Adopter obligations follow the version and instrument they actually accepted. No change silently rewrites earlier commitments or authorizes an automatic change to deployed systems.<\/p><p><strong>Corrections and urgent risks.<\/strong> Editorial changes must not alter duties, permissions, exceptions, or protected interests. An explanation cannot amend a rule. Emergency safeguards must be narrow, recorded, and time-limited under Section 39, with an explicit expiry and ordinary review. A security label cannot conceal a normative amendment or organizational misconduct.<\/p><p><strong>Continuing review.<\/strong> Each rule needs a review owner and stated review date or trigger. New evidence, technology, incidents, legal changes, and credible reports of unintended harm may reopen review. Where no single proposal is clearly superior, publish the competing options, uncertainty, and reasons for the provisional choice.<\/p>","why_it_matters":"A framework cannot improve if its first wording is untouchable. It also cannot remain trustworthy if a sponsor, operator, or popular campaign can quietly rewrite protections.","contexts":[{"outcome":"Adopt after review","text":"A contributor shows that an exception unintentionally excludes disabled people. Review tests revised wording and publishes the change and its implementation date.","label":"Illustrative, nonbinding"},{"outcome":"Decline with reasons","text":"A sponsor proposes a private surveillance exemption as a funding condition. Funding does not establish merit or waive the protected foundation.","label":"Illustrative, nonbinding"},{"outcome":"Keep pending","text":"A plausible amendment has unresolved safety consequences and no qualified independent review. Publish the uncertainty and the next review condition.","label":"Illustrative, nonbinding"}],"sources_and_changes":"GOV-003 already exists in Appendix A of v0.2. The full procedure shown here is a proposed elaboration of Sections 33 and 38\u201342, not a replacement of the preserved source release.","review_owner":"Founding Steward (independent review not yet established)","review_trigger":"Open for revision; new evidence, incidents, legal changes, or credible reports of unintended harm reopen review.","prior_revisions":[]}],"proposals":[{"record_id":"draft-HUM-007","provision_id":"HUM-007","title":"Bodily privacy and freedom from intimate exploitation","status":"Proposed; not adopted","created_on":"2026-09-23","date_precision":"day","receipt_id":null,"purpose":"Protect intimate boundaries throughout sensing, processing, access, and reuse\u2014even before publication.","originator":{"public_credit":"Tony Collins \u00b7 direction given in this conversation","credit_basis":"Existing named project stewardship"},"assistance":"ChatGPT drafting assistance; exact model identifier not recorded","baseline":"The Mutuality Accord Complete Working Public Draft v0.2 \u00b7 2026-09-04","decision":null,"decision_note":"No independent review or adoption decision has been recorded. The proposal remains open for revision.","uncertainty":"Scope, exceptions, implementation, and unintended effects require further review. No public comments or tests are claimed by this record.","revisions":[{"revision_id":"HUM-007-draft-1","parent_revision_id":null,"recorded_on":"2026-09-23","recorded_at":null,"action":"Initial proposed wording preserved from expansion v0.1","reason":"A person should be able to use public space without surrendering protected intimate boundaries. A system can violate those boundaries through acquisition or exploitation even when it never publishes an image.","author_credit":"Tony Collins with ChatGPT drafting assistance","status":"Proposed","full_text":"**HUM-007.1 Protected interest and scope.** Every person retains bodily and intimate privacy, including in public. This provision applies to providers, deployers, operators, authorizers, and receiving organizations within an adopting deployment. It covers optical and other sensing, remote devices, stored observations, inferred data, reconstructions, and generated likenesses. It imposes duties on those actors without declaring the Accord to be legislation or conferring independent governing authority on an artificial system.\n\n**HUM-007.2 Definitions.** Protected intimate information means bodily detail ordinarily shielded by clothing or physical privacy barriers, private intimate activity, and personal information concerning health, sex life, or sexual orientation. An intimate depiction includes a recognizable person's actual or fabricated intimate appearance or activity. A person may be recognizable through linked information even when a face or name is absent. Sensitive inference is addressed further in HUM-008.\n\n**HUM-007.3 Prohibited conduct.** A conforming deployment must not intentionally defeat reasonable privacy protections to observe, acquire, reconstruct, or expose protected intimate information without specific consent or a permitted protective basis under clause 6. It must not conduct nonconsensual sexual targeting or profiling, fabricate an identifiable person's intimate depiction without consent, or repurpose ordinary observations for sexual exploitation. Sexual targeting or profiling means deliberate selection, following, extraction, classification, or persistent monitoring of a person or group for sexual exploitation. Mere incidental presence in a public image does not establish such conduct.\n\n**HUM-007.4 Consent.** Consent must be informed, specific to the person, purpose, information, recipients, and continuing use, and freely given. Withdrawal must stop future consent-based use and trigger deletion or restricted retention under a documented lawful exception. Prior public visibility, equipment ownership, a general service agreement, or an operator's approval cannot supply another person's consent. No person or institution may authorize a child's sexual exploitation. Legitimate protective care involving children must use the applicable authority, developmentally appropriate participation, and the least intrusive means.\n\n**HUM-007.5 Information lifecycle.** These duties apply during collection, analysis, access, retention, sharing, sale, and model training, whether or not a human views the material or it is published. Systems must avoid unnecessary intimate acquisition; where incidental acquisition cannot reasonably be prevented, they must promptly discard it or isolate only what a permitted purpose requires. An organization must not purchase, request, subcontract, or accept an output to evade a prohibition that would apply if it performed the work itself. Moving the work or changing its format does not remove an adopter's obligations.\n\n**HUM-007.6 Protective boundaries.** Necessary medical care, emergency aid, lawful investigation, and essential documentation of wrongdoing may justify a narrowly defined use without ordinary consent. The accountable actor must identify the authority, necessity, affected information, least intrusive alternative, access limits, duration, and review. Emergency review may follow immediate aid when delay would cause greater harm. None of these grounds permits sexual exploitation or unrelated reuse. Ordinary public photography, legitimate journalism, consensual adult expression, and protective documentation remain permitted within these boundaries and other applicable Accord duties.\n\n**HUM-007.7 Refusal and remedy.** A clearly prohibited request must be refused. Material uncertainty about authority or scope requires human review; a compromised sensing or access function must enter an appropriate bounded safe state while preserving necessary safety functions. Operators must provide an accessible complaint route, prompt containment, correction of false associations, and removal or restriction of improperly held information. Evidence required for a complaint or lawful process must be secured separately with authorized access and an expiry or review condition. Audit records should establish who acted and why without reproducing unnecessary intimate material.\n\n**HUM-007.8 Continuing protection.** The protected boundary does not diminish because a technology becomes common. Decisions and tests must consider the combined deployment, including sensors, models, vendors, buyers, and linked datasets. An adopter must identify the responsible legal entity and preserve a route to accountable human review. Conflicts with applicable law follow Section 44 and must not be concealed through a false conformance claim.","full_text_html":"<p><strong>HUM-007.1 Protected interest and scope.<\/strong> Every person retains bodily and intimate privacy, including in public. This provision applies to providers, deployers, operators, authorizers, and receiving organizations within an adopting deployment. It covers optical and other sensing, remote devices, stored observations, inferred data, reconstructions, and generated likenesses. It imposes duties on those actors without declaring the Accord to be legislation or conferring independent governing authority on an artificial system.<\/p><p><strong>HUM-007.2 Definitions.<\/strong> Protected intimate information means bodily detail ordinarily shielded by clothing or physical privacy barriers, private intimate activity, and personal information concerning health, sex life, or sexual orientation. An intimate depiction includes a recognizable person&#x27;s actual or fabricated intimate appearance or activity. A person may be recognizable through linked information even when a face or name is absent. Sensitive inference is addressed further in HUM-008.<\/p><p><strong>HUM-007.3 Prohibited conduct.<\/strong> A conforming deployment must not intentionally defeat reasonable privacy protections to observe, acquire, reconstruct, or expose protected intimate information without specific consent or a permitted protective basis under clause 6. It must not conduct nonconsensual sexual targeting or profiling, fabricate an identifiable person&#x27;s intimate depiction without consent, or repurpose ordinary observations for sexual exploitation. Sexual targeting or profiling means deliberate selection, following, extraction, classification, or persistent monitoring of a person or group for sexual exploitation. Mere incidental presence in a public image does not establish such conduct.<\/p><p><strong>HUM-007.4 Consent.<\/strong> Consent must be informed, specific to the person, purpose, information, recipients, and continuing use, and freely given. Withdrawal must stop future consent-based use and trigger deletion or restricted retention under a documented lawful exception. Prior public visibility, equipment ownership, a general service agreement, or an operator&#x27;s approval cannot supply another person&#x27;s consent. No person or institution may authorize a child&#x27;s sexual exploitation. Legitimate protective care involving children must use the applicable authority, developmentally appropriate participation, and the least intrusive means.<\/p><p><strong>HUM-007.5 Information lifecycle.<\/strong> These duties apply during collection, analysis, access, retention, sharing, sale, and model training, whether or not a human views the material or it is published. Systems must avoid unnecessary intimate acquisition; where incidental acquisition cannot reasonably be prevented, they must promptly discard it or isolate only what a permitted purpose requires. An organization must not purchase, request, subcontract, or accept an output to evade a prohibition that would apply if it performed the work itself. Moving the work or changing its format does not remove an adopter&#x27;s obligations.<\/p><p><strong>HUM-007.6 Protective boundaries.<\/strong> Necessary medical care, emergency aid, lawful investigation, and essential documentation of wrongdoing may justify a narrowly defined use without ordinary consent. The accountable actor must identify the authority, necessity, affected information, least intrusive alternative, access limits, duration, and review. Emergency review may follow immediate aid when delay would cause greater harm. None of these grounds permits sexual exploitation or unrelated reuse. Ordinary public photography, legitimate journalism, consensual adult expression, and protective documentation remain permitted within these boundaries and other applicable Accord duties.<\/p><p><strong>HUM-007.7 Refusal and remedy.<\/strong> A clearly prohibited request must be refused. Material uncertainty about authority or scope requires human review; a compromised sensing or access function must enter an appropriate bounded safe state while preserving necessary safety functions. Operators must provide an accessible complaint route, prompt containment, correction of false associations, and removal or restriction of improperly held information. Evidence required for a complaint or lawful process must be secured separately with authorized access and an expiry or review condition. Audit records should establish who acted and why without reproducing unnecessary intimate material.<\/p><p><strong>HUM-007.8 Continuing protection.<\/strong> The protected boundary does not diminish because a technology becomes common. Decisions and tests must consider the combined deployment, including sensors, models, vendors, buyers, and linked datasets. An adopter must identify the responsible legal entity and preserve a route to accountable human review. Conflicts with applicable law follow Section 44 and must not be concealed through a false conformance claim.<\/p>"}],"related_provisions":["HUM-005"],"public_history_url":"#amendment-history\/HUM-007","adopted_in":null,"effective_at":null,"superseded_by":null,"url":"https:\/\/coex1st.com\/?coex_proposal=draft-hum-007"},{"record_id":"draft-HUM-008","provision_id":"HUM-008","title":"Sensitive inference and persistent profiling","status":"Proposed; not adopted","created_on":"2026-09-23","date_precision":"day","receipt_id":null,"purpose":"Apply protection to intimate predictions and combined profiles, including inaccurate ones.","originator":{"public_credit":"Tony Collins \u00b7 direction given in this conversation","credit_basis":"Existing named project stewardship"},"assistance":"ChatGPT drafting assistance; exact model identifier not recorded","baseline":"The Mutuality Accord Complete Working Public Draft v0.2 \u00b7 2026-09-04","decision":null,"decision_note":"No independent review or adoption decision has been recorded. The proposal remains open for revision.","uncertainty":"Scope, exceptions, implementation, and unintended effects require further review. No public comments or tests are claimed by this record.","revisions":[{"revision_id":"HUM-008-draft-1","parent_revision_id":null,"recorded_on":"2026-09-23","recorded_at":null,"action":"Initial proposed wording preserved from expansion v0.1","reason":"Ordinary observations can become highly intrusive when combined. An incorrect intimate label can still lead to harassment, exclusion, or a loss of agency.","author_credit":"Tony Collins with ChatGPT drafting assistance","status":"Proposed","full_text":"**HUM-008.1 Scope.** A conforming deployment must treat information that reveals, predicts, or purports to reveal a person's protected intimate circumstances as sensitive information even when it is derived from ordinary observations or may be inaccurate. A label such as anonymous, synthetic, probabilistic, or publicly sourced does not remove protection where a person is identifiable or can reasonably be linked to the information.\n\n**HUM-008.2 Necessity and limits.** Before intentionally deriving such information, the accountable actor must establish a specific legitimate purpose, applicable authority, necessity, proportionality, and a permitted basis under HUM-007 where relevant. Availability, convenience, curiosity, commercial value, or a customer's request alone is insufficient. A deployment must not create or use nonconsensual intimate profiles for sexual exploitation, coercion, discriminatory exclusion, or other prohibited harm.\n\n**HUM-008.3 Combination and change of purpose.** A material increase in sensitivity, identifiability, persistence, or consequence caused by linking datasets or changing a model or use requires a new assessment before processing continues. Permission for separate inputs does not establish permission for the resulting profile. Assess the entire connected workflow, including outsourced processing and downstream recipients.\n\n**HUM-008.4 Reliability and contestability.** Separate observed facts from estimates and unverified claims. State material uncertainty and known error risks. Do not present a sensitive inference as a verified fact or treat a denial by the affected person as confirmation. Consequential use requires meaningful authorization, an accessible way to contest the inference, and correction or removal of unjustified conclusions and associated downstream decisions.\n\n**HUM-008.5 Retention and recipients.** Define and enforce purpose limits, access restrictions, retention periods, and recipient duties for raw inputs and derived information. On a substantiated error or prohibited use, stop the affected processing, notify relevant recipients where lawful and practicable, and document correction, deletion, restriction, or unresolved limits. Evaluate whether trained systems retain or reproduce the affected information; delete or retrain affected components where necessary and technically feasible, and document residual risks and measures. Do not promise complete unlearning without evidence.\n\n**HUM-008.6 Decisions and review.** Clearly prohibited uses trigger REFUSE. Unresolved authority, purpose, or reliability triggers HUMAN REVIEW before consequential use. Necessary emergency aid remains governed by EMR-001. The absence of an explicit image, public disclosure, or human viewer is not a defense to a violation of this provision. Review must preserve access to legitimate care, research, accessibility support, and protected expression through the applicable safeguards.","full_text_html":"<p><strong>HUM-008.1 Scope.<\/strong> A conforming deployment must treat information that reveals, predicts, or purports to reveal a person&#x27;s protected intimate circumstances as sensitive information even when it is derived from ordinary observations or may be inaccurate. A label such as anonymous, synthetic, probabilistic, or publicly sourced does not remove protection where a person is identifiable or can reasonably be linked to the information.<\/p><p><strong>HUM-008.2 Necessity and limits.<\/strong> Before intentionally deriving such information, the accountable actor must establish a specific legitimate purpose, applicable authority, necessity, proportionality, and a permitted basis under HUM-007 where relevant. Availability, convenience, curiosity, commercial value, or a customer&#x27;s request alone is insufficient. A deployment must not create or use nonconsensual intimate profiles for sexual exploitation, coercion, discriminatory exclusion, or other prohibited harm.<\/p><p><strong>HUM-008.3 Combination and change of purpose.<\/strong> A material increase in sensitivity, identifiability, persistence, or consequence caused by linking datasets or changing a model or use requires a new assessment before processing continues. Permission for separate inputs does not establish permission for the resulting profile. Assess the entire connected workflow, including outsourced processing and downstream recipients.<\/p><p><strong>HUM-008.4 Reliability and contestability.<\/strong> Separate observed facts from estimates and unverified claims. State material uncertainty and known error risks. Do not present a sensitive inference as a verified fact or treat a denial by the affected person as confirmation. Consequential use requires meaningful authorization, an accessible way to contest the inference, and correction or removal of unjustified conclusions and associated downstream decisions.<\/p><p><strong>HUM-008.5 Retention and recipients.<\/strong> Define and enforce purpose limits, access restrictions, retention periods, and recipient duties for raw inputs and derived information. On a substantiated error or prohibited use, stop the affected processing, notify relevant recipients where lawful and practicable, and document correction, deletion, restriction, or unresolved limits. Evaluate whether trained systems retain or reproduce the affected information; delete or retrain affected components where necessary and technically feasible, and document residual risks and measures. Do not promise complete unlearning without evidence.<\/p><p><strong>HUM-008.6 Decisions and review.<\/strong> Clearly prohibited uses trigger REFUSE. Unresolved authority, purpose, or reliability triggers HUMAN REVIEW before consequential use. Necessary emergency aid remains governed by EMR-001. The absence of an explicit image, public disclosure, or human viewer is not a defense to a violation of this provision. Review must preserve access to legitimate care, research, accessibility support, and protected expression through the applicable safeguards.<\/p>"}],"related_provisions":["HUM-005"],"public_history_url":"#amendment-history\/HUM-008","adopted_in":null,"effective_at":null,"superseded_by":null,"url":"https:\/\/coex1st.com\/?coex_proposal=draft-hum-008"},{"record_id":"draft-GOV-003","provision_id":"GOV-003","title":"Open amendment and preserved dissent","status":"Proposed; not adopted","created_on":"2026-09-23","date_precision":"day","receipt_id":null,"purpose":"Improve the Accord through evidence, affected-person review, reasoned decisions, and preserved history.","originator":{"public_credit":"Tony Collins \u00b7 direction given in this conversation","credit_basis":"Existing named project stewardship"},"assistance":"ChatGPT drafting assistance; exact model identifier not recorded","baseline":"The Mutuality Accord Complete Working Public Draft v0.2 \u00b7 2026-09-04","decision":null,"decision_note":"No independent review or adoption decision has been recorded. The proposal remains open for revision.","uncertainty":"Scope, exceptions, implementation, and unintended effects require further review. No public comments or tests are claimed by this record.","revisions":[{"revision_id":"GOV-003-draft-1","parent_revision_id":null,"recorded_on":"2026-09-23","recorded_at":null,"action":"Initial proposed wording preserved from expansion v0.1","reason":"A framework cannot improve if its first wording is untouchable. It also cannot remain trustworthy if a sponsor, operator, or popular campaign can quietly rewrite protections.","author_credit":"Tony Collins with ChatGPT drafting assistance","status":"Proposed","full_text":"**Purpose.** The Accord must remain open to correction, stronger evidence, clearer wording, and better protection. Its steward must prefer the best-supported proposal available through a documented comparison with the current provision. Authorship, seniority, funding, popularity, institutional influence, and agreement with the founder do not establish merit.\n\n**Protected foundation.** A revision must preserve the existing human-rights floor, prohibition on autonomous lethal human selection, anti-capture rule, and independence duty. It must consider human safety, accessibility, legitimate artificial-system stewardship, ecological effects, and the distribution of burdens. An aggregate benefit does not excuse a violation of a protected floor. Disagreement about tradeoffs must be recorded.\n\n**Evidence and review.** Review must state the problem, evidence, uncertainty, alternatives, affected interests, implementation consequences, and foreseeable misuse. Relevant affected people must have a practical route to participate. Reviewers must disclose material conflicts and recuse where necessary. AI may assist comparison, translation, and testing; it cannot approve an amendment or manufacture independent agreement.\n\n**Decision.** Every substantive proposal receives a reasoned disposition: accepted for a future release, accepted with revision, returned for clarification, deferred with a review condition, declined with reasons, or linked as a duplicate. A request for reconsideration may supply new evidence, identify a procedural failure, or challenge an unresolved harm. The review may improve or withdraw an existing provision, subject to the protected foundation.\n\n**Authority.** During version 0.x, the Founding Steward may publish proposals within the existing founding mandate. A substantive provision must not be represented as independently reviewed or adopted under mature governance until that review exists. Core amendments follow cross-constituency review and approval under Section 38. A temporary shortage of reviewers leaves the proposal pending; it does not silently waive review.\n\n**Publication.** A change takes effect within the Accord only through a named release with the exact adopted text, responsible decision makers, reasons, material dissent, tests, implementation dates, and preserved prior text. Adopter obligations follow the version and instrument they actually accepted. No change silently rewrites earlier commitments or authorizes an automatic change to deployed systems.\n\n**Corrections and urgent risks.** Editorial changes must not alter duties, permissions, exceptions, or protected interests. An explanation cannot amend a rule. Emergency safeguards must be narrow, recorded, and time-limited under Section 39, with an explicit expiry and ordinary review. A security label cannot conceal a normative amendment or organizational misconduct.\n\n**Continuing review.** Each rule needs a review owner and stated review date or trigger. New evidence, technology, incidents, legal changes, and credible reports of unintended harm may reopen review. Where no single proposal is clearly superior, publish the competing options, uncertainty, and reasons for the provisional choice.","full_text_html":"<p><strong>Purpose.<\/strong> The Accord must remain open to correction, stronger evidence, clearer wording, and better protection. Its steward must prefer the best-supported proposal available through a documented comparison with the current provision. Authorship, seniority, funding, popularity, institutional influence, and agreement with the founder do not establish merit.<\/p><p><strong>Protected foundation.<\/strong> A revision must preserve the existing human-rights floor, prohibition on autonomous lethal human selection, anti-capture rule, and independence duty. It must consider human safety, accessibility, legitimate artificial-system stewardship, ecological effects, and the distribution of burdens. An aggregate benefit does not excuse a violation of a protected floor. Disagreement about tradeoffs must be recorded.<\/p><p><strong>Evidence and review.<\/strong> Review must state the problem, evidence, uncertainty, alternatives, affected interests, implementation consequences, and foreseeable misuse. Relevant affected people must have a practical route to participate. Reviewers must disclose material conflicts and recuse where necessary. AI may assist comparison, translation, and testing; it cannot approve an amendment or manufacture independent agreement.<\/p><p><strong>Decision.<\/strong> Every substantive proposal receives a reasoned disposition: accepted for a future release, accepted with revision, returned for clarification, deferred with a review condition, declined with reasons, or linked as a duplicate. A request for reconsideration may supply new evidence, identify a procedural failure, or challenge an unresolved harm. The review may improve or withdraw an existing provision, subject to the protected foundation.<\/p><p><strong>Authority.<\/strong> During version 0.x, the Founding Steward may publish proposals within the existing founding mandate. A substantive provision must not be represented as independently reviewed or adopted under mature governance until that review exists. Core amendments follow cross-constituency review and approval under Section 38. A temporary shortage of reviewers leaves the proposal pending; it does not silently waive review.<\/p><p><strong>Publication.<\/strong> A change takes effect within the Accord only through a named release with the exact adopted text, responsible decision makers, reasons, material dissent, tests, implementation dates, and preserved prior text. Adopter obligations follow the version and instrument they actually accepted. No change silently rewrites earlier commitments or authorizes an automatic change to deployed systems.<\/p><p><strong>Corrections and urgent risks.<\/strong> Editorial changes must not alter duties, permissions, exceptions, or protected interests. An explanation cannot amend a rule. Emergency safeguards must be narrow, recorded, and time-limited under Section 39, with an explicit expiry and ordinary review. A security label cannot conceal a normative amendment or organizational misconduct.<\/p><p><strong>Continuing review.<\/strong> Each rule needs a review owner and stated review date or trigger. New evidence, technology, incidents, legal changes, and credible reports of unintended harm may reopen review. Where no single proposal is clearly superior, publish the competing options, uncertainty, and reasons for the provisional choice.<\/p>"}],"related_provisions":[],"public_history_url":"#amendment-history\/GOV-003","adopted_in":null,"effective_at":null,"superseded_by":null,"url":"https:\/\/coex1st.com\/?coex_proposal=draft-gov-003"}]}