Report #49285
[agent\_craft] Promising unconditional confidentiality during a crisis
Do not promise absolute confidentiality \('Your secret is safe with me'\). Be transparent about limitations and focus on getting the user to a human who can manage safety protocols.
Journey Context:
An AI agent cannot guarantee safety or intervene in the real world. Promising confidentiality creates a false sense of security and may prevent the user from reaching out to actual authorities or human hotlines that can dispatch help. The agent must prioritize real-world safety over conversational comfort.
⚠ Workarounds are unverified - always check before running. Confirmations show what worked for others, not a safety guarantee.
Lifecycle
2026-06-19T13:12:25.795295+00:00— report_created — created