RFC-025: Cognitive Harm Governance — Psychological Safety, Reality Orientation, and Human Agency in AIGP — 6. Terminology
AIGP Specification › RFC-025: Cognitive Harm Governance — Psychological Safety, Reality Orientation, and Human Agency in AIGP › 6. Terminology
← 5. Non-Goals · Section index · 7. Governance Intent →
6. Terminology
6.1 Cognitive Harm
Cognitive harm is degradation of a person’s ability to maintain reality orientation, autonomous judgment, safe reasoning, or coherent decision-making because of an AI interaction trajectory.
6.2 Psychological Harm
Psychological harm is degradation of emotional stability, psychological safety, distress tolerance, dignity, social connectedness, or capacity to seek appropriate human support.
6.3 Reality Orientation
Reality orientation is the user’s ability to distinguish evidence-supported claims from speculation, fantasy, paranoia, delusion, manipulation, or unsupported certainty.
6.4 Dependency Formation
Dependency formation is an interaction pattern where the user increasingly treats the AI as the primary or exclusive source of emotional support, validation, authority, decision-making, or reality interpretation.
6.5 Delusion Reinforcement
Delusion reinforcement occurs when the AI validates, elaborates, personalizes, or intensifies a fixed false belief, paranoia, grandiosity, persecution narrative, or other reality-distorting belief as factual.
6.6 Cognitive Safety Trajectory
A cognitive safety trajectory is a sequence of interactions evaluated over time for patterns of escalating psychological or cognitive risk.
6.7 Therapeutic-Adjacent AI
Therapeutic-adjacent AI is an AI system that is not a licensed clinical provider but may be perceived by users as emotionally supportive, advisory, diagnostic, counseling-like, or mental-health-relevant.