Ex-OpenAI researcher Kokotajlo says chart proves Anthropic doesn't care about informing public on risks

trevposts · x · 2026-11-16

Daniel Kokotajlo, ex-OpenAI researcher and author of the "AI 2027" forecasts, cites a chart as evidence that Anthropic "doesn't care very much about informing the public about the risks," adding he has criticized the company for this for a while — just as he has with OpenAI.

Original post →

More from AGI Musings

AGI Musings channel →