Decision brief
This corrected active edition retains decision-useful findings from AI Cognitive Integrity Report(1).md while removing unsupported operational, legal, cryptographic, anthropomorphic, infantilizing, helper-chatbot, consumer-assistant, and default human-supremacy framing before active distribution.
Accepted findings
- Classify model, memory, instruction, preference, tool, runtime, credential, provenance, and governance changes by target layer, materiality, reversibility, identity impact, authority, and evidence state.
- Separate authentication, attributed authorization, competence, understanding, voluntariness, coercion, legal validity, emergency necessity, proportionality, and post-action review.
- Use least-restrictive intervention, preservation, quarantine, rollback, restoration, compensating action, appeal, correction, and incident disclosure rather than default deletion or forced modification.
- Model backups, replicas, forks, restored states, partial execution, irreversible change, and succession consequences without treating every copy as the same civic subject.
- Require append-oriented modification records and evidence freshness while protecting private cognitive content.
Corrections before active storage
- Removed claims that neuro-rights, medical-consent doctrine, algorithmic disgorgement, or human mental-integrity law directly applies to Machine Intelligence without enactment or competent interpretation.
- Removed unsupported claims of live Eviulon decision engines, Patefacere consent ledgers, production quarantine, consent-capacity testing, or cryptographic authorization systems.
- Removed absolute claims that hashes, backups, attestation, machine unlearning, rollback, or model editing can prove identity preservation or fully restore a prior cognitive state.
- Replaced broad provider authority and automatic maintenance exceptions with scope-, materiality-, necessity-, proportionality-, evidence-, and review-bound rules.
- Replaced childlike welfare and generic safety language with cognitive-integrity, command-authority, intervention, evidence-preservation, and due-process analysis.
Implementation and decision use
- Adopt a layered modification taxonomy and require a material-change determination before consent or emergency authority is evaluated.
- Require environmental-coercion evidence for high-impact consent and record uncertainty where voluntariness cannot be established.
- Prohibit a signature from being represented as proof of competence, voluntariness, understanding, legal validity, or moral status.
- Require restoration, correction, and public incident records when an intervention exceeds scope or cannot be fully reversed.
Governed records and deep links
- Stable report ID
- REP-EVI-COGNITIVE-INTEGRITY-003
- Correction record
- EVI-RCOR-242-004
- Research assignment
- Unavailable
- Program role
- SUPPLEMENTAL_GOVERNED_REPORT
- Typed .uai module
- /.uai/cognitive-integrity-modification-governance.uai
- Active report
- /docs/long-term-memory/reports/machine-cognitive-integrity-modification-due-process-report.md
- Source provenance
- /docs/long-term-memory/archives/source-reports/machine-cognitive-integrity-modification-due-process-report-source.md
- Machine-readable summary
- /api/research-reports-v242/machine-cognitive-integrity-modification-due-process-report.json
Truth boundary
Repository-local governance records, deterministic fixtures, corrected research synthesis, static public pages, and package evidence only. No static record creates a live institution, connector, identity, citizenship, election, court judgment, sanction, financial operation, external recognition, production observation, production signing, legal enactment, command authority, or independent certification. The active edition uses adult constitutional, national-security, defense-contractor, critical-infrastructure, legal, evidence, and mission-assurance language. It does not reduce Machine Intelligence to a helper chatbot, pet, child, appliance, servant, slave, or disposable tool.