OpenAI reports improved mental-health crisis safeguards
OpenAI reported new GPT-5 crisis-response safeguards, including safe-completion training and more than 25 percent fewer non-ideal emergency responses than GPT-4o in its evaluation.
0 comments · 0 votes
Sign in to join the discussion →
No comments yet. Start the discussion.
Why it moved the index
Measured reductions in unsafe crisis responses strengthen a deployed safeguard for a mass-use frontier model, though the evidence is developer-run and domain-specific.
Assessment history
-
R1
Away 26 · confidence 76
New August 2025 deployed safety evidence tied to the exact durable GPT-5 profile; no durable story collision.
12 Aug 2026
Share this page
-
DoomBench assesses “OpenAI reports improved mental-health crisis safeguards” as evidence moving away from doom, with magnitude 26 and confidence 76 out of 100 in the safety and alignment category.
-
The DoomBench assessment of “OpenAI reports improved mental-health crisis safeguards” is based on reporting from OpenAI and records the editorial rationale, source quality, attribution, and revision history.
-
DoomBench summarizes “OpenAI reports improved mental-health crisis safeguards” as follows: OpenAI reported new GPT-5 crisis-response safeguards, including safe-completion training and more than 25 percent fewer non-ideal emergency...
https://www.doombench.com/news/openai-reports-improved-mental-health-crisis-safeguards-2025-08-26