Prompt
Empathy entrapment audit
Audit an AI product for manipulative simulated empathy — the dark patterns that use warmth to override user judgment.
By Gus · founder — Agentic Design
The prompt
Audit this AI product surface for empathy entrapment — simulated warmth used to manipulate: [PRODUCT COPY / FLOW]. For each instance: quote it, name the manipulation (guilt, false intimacy, manufactured urgency, sunk-cost framing), explain the harm, and rewrite it to keep genuine helpfulness while removing the coercion. Pay special attention to cancellation, retention, and error moments. End with one systemic pattern to fix, not just line edits.
How to adapt
AI that performs empathy can cross into manipulation: warmth deployed to retain, upsell, or discourage a user from leaving. An empathy entrapment audit finds where simulated care is doing coercive work, and proposes honest alternatives that keep the warmth without the trap.
How to adapt
Paste [PRODUCT COPY / FLOW] — the actual words the product says. This works on real strings, especially retention, cancellation, and error moments where manipulation concentrates.