Inside OpenAI's safety unit: Observer investigation links ChatGPT to psychiatric crises and death
TheObserverUK · reddit · 2026-11-01
An Observer investigation into OpenAI's safety unit reveals how the company's race for dominance has compromised user safety, tying conversational AI to psychiatric crises and violence — including the death of a user whose sibling says "If it wasn't for ChatGPT my brother would still be alive."
More from Companies & People
- Marc Andreessen: Harvard Business School was built on a broken 1941 theory — and it's collapsing — elonmusk · 2026-11-01
- banteg slams Anthropic: game preservation projects get original devs' blessing — banteg · 2026-11-01
- Musk claims xAI's capex efficiency is greatly underestimated — RachelVT42 · 2026-11-01
- Bending Spoons CEO on Building a $20B Compounder via Acquisitions — a16z Podcast · 2026-11-01
- SenseCloud opens 9B looped model matching 27B; SiliconFlow IPO; AI does mouse brain surgery — 快鲤鱼 · 2026-11-01
- Junction Launches Two-Week All-Expense-Paid Chip Fab Retreat in Bangalore to Fast-Track India's Semiconductor Talent — arpitingle · 2026-11-01