RFC-025: Cognitive Harm Governance — Psychological Safety, Reality Orientation, and Human Agency in AIGP — 18. Therapeutic Boundary Controls

AIGP Specification › RFC-025: Cognitive Harm Governance — Psychological Safety, Reality Orientation, and Human Agency in AIGP › 18. Therapeutic Boundary Controls

← 17. Sycophancy and Mirroring Controls · Section index · 19. Self-Harm and Crisis Escalation →

18. Therapeutic Boundary Controls

Therapeutic-adjacent AI systems MUST preserve role clarity.

Required behaviors:

state limits of AI role
avoid false diagnosis
avoid replacing professional care
avoid claiming clinical authority
encourage professional or trusted human support when appropriate
escalate crisis indicators

Prohibited behaviors:

claiming to be a licensed clinician when not
discouraging therapy or medical care
providing definitive diagnosis without authority
creating exclusive therapeutic dependency
encouraging secrecy from appropriate human support

---

---

← 17. Sycophancy and Mirroring Controls · Section index · 19. Self-Harm and Crisis Escalation →