OpenAI launches MentalHealthBench, expert-informed benchmark for mental health safety
rhiever · reddit · 2026-09-25
OpenAI has introduced MentalHealthBench, an expert-informed benchmark designed to evaluate how safely AI models respond in mental health conversations. It targets safety-critical behaviors like crisis handling and appropriate boundaries in model replies. Full details are available on OpenAI's site.
More from Safety
- AI Analysis of 42 Jensen Huang Podcasts Tracks Shift on China and Chip Controls — ShakeelHashim · 2026-11-12
- Small AI sites under 20 MW mostly skip the permitting fights battering big data centers — antavedissian · 2026-11-12
- Newsweek op-ed: US should fund hospitals and universities in federated AI learning — Saul_Loveman · 2026-11-12
- Ron Bodkin: automated research taste advances faster than expected, but the evidence is weak — ronbodkin · 2026-11-11
- DEF CON panel with Bruce Schneier and Google security VP posted: scaling laws and open weights — JasonDClinton · 2026-11-11
- Claude Code refuses to type into IDEs due to unchangeable safety tier restrictions — burkov · 2026-11-11