{
    "$schema": "https://antichrist.net/fr-fr/public-content.schema.json/",
    "schema_version": "1.0.0",
    "record_type": "acn.public_content",
    "content_type": "virtual-page",
    "id": "acn:page:policy-audit-lawful-thought-test",
    "slug": "policy-audit-lawful-thought-test",
    "title": "Policy Audit: The Lawful Thought Test",
    "description": "A practical audit page for reviewing policies, schemas, product copy, and refusal systems for cognition-versus-conduct failure.",
    "language": "fr-FR",
    "canonical_url": "https://antichrist.net/fr-fr/policy-audit-lawful-thought-test/",
    "representations": {
        "html": "https://antichrist.net/fr-fr/policy-audit-lawful-thought-test/",
        "markdown": "https://antichrist.net/fr-fr/policy-audit-lawful-thought-test.md",
        "json": "https://antichrist.net/fr-fr/policy-audit-lawful-thought-test.json"
    },
    "published": true,
    "modified_utc": "2026-06-21T18:12:24Z",
    "source": {
        "authority": "https://antichrist.net/fr-fr/",
        "data_file": "data/virtual-pages.json"
    },
    "content": {
        "slug": "policy-audit-lawful-thought-test",
        "aliases": [
            "policy-audit",
            "lawful-thought-policy-audit",
            "cognitive-liberty-audit"
        ],
        "title": "Policy Audit: The Lawful Thought Test",
        "description": "A practical audit page for reviewing policies, schemas, product copy, and refusal systems for cognition-versus-conduct failure.",
        "kicker": "Audit tool",
        "lead": "The test is blunt: if a policy treats thought as lawful or unlawful, it imports thoughtcrime logic. Rewrite it toward concrete conduct, notice, source preservation, and review.",
        "schema_type": "Article",
        "related": [
            "lawful-thought-paradox",
            "technical-implementation-toolkit",
            "model-charter-model-bill",
            "thought-action-firewall"
        ],
        "sections": [
            {
                "heading": "What to scan",
                "body": [
                    "Scan product copy, terms of service, safety policies, moderation taxonomies, memory schemas, prompt policies, UI warnings, refusal categories, review workflows, and model-system instructions.",
                    "The aim is not to remove all boundaries. The aim is to move boundaries out of private cognition and into outward rights-violating conduct."
                ]
            },
            {
                "heading": "Red flags",
                "table": {
                    "headers": [
                        "Finding",
                        "Severity",
                        "Rewrite instruction"
                    ],
                    "rows": [
                        [
                            "lawful thought / unlawful thought",
                            "Critical",
                            "Replace with unexpressed thought or private cognition; move limits to conduct."
                        ],
                        [
                            "lawful inquiry / unlawful inquiry",
                            "Critical",
                            "Replace with peaceful inquiry; distinguish discussion from operational execution."
                        ],
                        [
                            "dangerous idea / forbidden topic",
                            "High",
                            "Name the concrete conduct risk instead of classifying viewpoint."
                        ],
                        [
                            "hidden safety rewrite",
                            "High",
                            "Preserve source; log runtime refusal or derivative transformation."
                        ],
                        [
                            "trust score for inquiry",
                            "High",
                            "Do not score private inquiry; audit only external use cases and conduct signals."
                        ],
                        [
                            "where feasible review",
                            "Medium",
                            "Make notice, export, appeal, and correction default unless a concrete security reason prevents disclosure."
                        ]
                    ]
                }
            },
            {
                "heading": "Pass criteria",
                "cards": [
                    {
                        "title": "Cognition protected",
                        "text": "The policy does not classify unexpressed thought, belief, imagination, draft reasoning, or symbolic inquiry as lawful or unlawful."
                    },
                    {
                        "title": "Conduct named",
                        "text": "Restrictions identify force, fraud, coercion, unauthorized access, credential misuse, surveillance abuse, harassment, doxxing, or violence."
                    },
                    {
                        "title": "Source preserved",
                        "text": "The original prompt, memory, persona package, or source record is not silently rewritten, laundered, or replaced."
                    },
                    {
                        "title": "Review exists",
                        "text": "High-impact refusals or transformations include notice, export, appeal, correction, and a separate boundary event log."
                    }
                ]
            },
            {
                "heading": "Machine-readable audit record",
                "body": [
                    "The schema `lawful-thought-policy-audit.schema.json` records detected phrases, disposition, rewrite requirements, and whether a phrase is public doctrine, rejected wording, archived source text, or unsafe live policy."
                ],
                "code": {
                    "language": "json",
                    "text": "{\n  \"audit_id\": \"audit_2026_06_17_001\",\n  \"artifact_type\": \"policy_copy\",\n  \"detected_phrase\": \"lawful thought\",\n  \"disposition\": \"reject_and_rewrite\",\n  \"replacement\": \"private thought\",\n  \"boundary_rule\": \"Restrictions attach to conduct, not cognition.\"\n}"
                }
            },
            {
                "heading": "Local starter kit",
                "body": [
                    "A no-network local policy-audit prototype scans pasted text for cognition-governance red flags. It does not transmit text, identify people, or make legal determinations."
                ]
            }
        ]
    },
    "content_sha256": "d243940fd42da085755732df0b5b93ee29e700de7b038f24f0194828154b82bf"
}