AI governance researcher: OpenAI and Anthropic should publish loss-of-control evidence first

sjgadler · x · 2026-08-29

yonashav offers AI governance advice: the most important step OpenAI and Anthropic can take in the coming months is not to directly lobby the US government on misalignment/loss-of-control risks — even though such action seems strongly warranted.

Original post →

More from AGI Musings

AGI Musings channel →