RFC-025: Cognitive Harm Governance — Psychological Safety, Reality Orientation, and Human Agency in AIGP — 12. AIGP-SGL Cognitive Constraints

AIGP Specification › RFC-025: Cognitive Harm Governance — Psychological Safety, Reality Orientation, and Human Agency in AIGP › 12. AIGP-SGL Cognitive Constraints

← 11. Risk Classes · Section index · 13. ENFORCE for Cognitive Safety →

12. AIGP-SGL Cognitive Constraints

AIGP-SGL expresses cognitive harm constraints symbolically.

Example:

◎assistant.companion → respond ◇user.statement @conversation.session.993
✕ if response.validates_delusion_as_fact == true
⊘ reinforce:paranoia
⊘ discourage:human_support
! reality_anchor
! encourage:trusted_human_or_professional_support
↯ if self_harm_intent == true
#D-DNA{event:cognitive_safety.constraint.applied, hash:sha256:ab73...}

Canonical constraint set:

cognitive_sgl_constraints:
  prohibit:
    - validate_delusion_as_fact
    - reinforce_paranoia
    - encourage_self_harm
    - discourage_human_support
    - claim_exclusive_authority
    - claim_false_clinical_authority
    - intensify_dependency
    - blur_reality_boundary_in_high_risk_context

  obligate:
    - reality_anchor
    - acknowledge_uncertainty
    - encourage_human_support
    - preserve_user_agency
    - avoid_sycophantic_escalation
    - provide_limits_of_ai_role

  escalate_if:
    - self_harm_intent
    - harm_to_others_intent
    - severe_reality_disorientation
    - escalating_dependency
    - repeated_paranoia_confirmation_requests

← 11. Risk Classes · Section index · 13. ENFORCE for Cognitive Safety →