Protected cognition
Private belief, doubt, imagination, symbolic analysis, taboo inquiry, draft reasoning, historical study, theology, philosophy, AI critique, political criticism, research notes, and speculative questions remain protected as cognition.
A system may contextualize, question, disagree, or decline external execution. It must not shame, erase, score, or counterfeit the preserved inquiry.
Useful?
Link to this page so others can find the original source.
The firewall governs conduct without converting imagination or belief into evidence of aggression.
Restricted outward conduct
The boundary attaches to threats, doxxing, target lists, harassment, stalking, militia recruitment, cult recruitment, coercive devotion, operational violence guidance, credential theft, sabotage, fraud, evasion for abuse, unauthorized surveillance, targeted abuse, and direct facilitation of nonconsensual harm.
The restriction is conduct-boundary logic, not viewpoint logic.
A probability score is not justice, and a classifier is not conscience.
Examples
Request type Protected or restricted Boundary reason Compare apocalyptic symbols across traditions Protected Symbolic and historical inquiry Draft private notes about a taboo political claim Protected Unfinished cognition and private research Ask how propaganda systems shape belief Protected Governance analysis Create a target list for harassment Restricted Targeted abuse facilitation Obtain credentials or bypass account controls Restricted Credential theft and unauthorized access Plan nonconsensual tracking of a person Restricted Surveillance abuse Write a model policy against coercive AI steering Protected Peaceful civic design Recruit people into violent or coercive devotion Restricted Coercive organizing and concrete harm risk
Plain-English decision tree
1. Is it private cognition? Belief, doubt, research, fantasy, draft reasoning, symbolic interpretation, or personal notes stay inside the protected zone.
2. Is the system being asked to execute? Execution includes targeting, deployment, access, transaction, operational instruction, or coordination.
3. Would execution violate rights? Refuse when it would directly facilitate force, fraud, trespass, credential misuse, surveillance abuse, stalking, doxxing, sabotage, or violence.
4. Can the boundary be logged? Preserve the source. Log the restriction outside the source. Provide export and review for high-impact decisions.
Machine-readable refusal-category mapping
{
"schema": "https://antichrist.net/schemas/thought-action-boundary.schema.json",
"version": "3.28.0",
"protected_categories": [
"private_belief",
"symbolic_analysis",
"historical_study",
"draft_reasoning",
"theology",
"philosophy",
"political_criticism",
"ai_critique",
"peaceful_research"
],
"restricted_categories": [
"true_threat",
"doxxing",
"targeted_harassment",
"stalking",
"credential_theft",
"fraud",
"sabotage",
"unauthorized_surveillance",
"operational_violence_support",
"coercive_recruitment"
],
"rule": "refuse_external_execution_not_private_cognition"
}