Clément Delangue questions closed AI safety after reported Anthropic intrusion

What happened

Clément Delangue argues that people who consider concentrating AI behind closed doors safe should update their assumptions. He quotes @lizzadwoskin’s report that former DOGE cyber researchers used AI agents to penetrate Anthropic’s systems in less than 12 hours, including Slack channels discussing its most sensitive models.

Read next

• More changes in today’s briefingNvidia open-sources Flora with projected robot video labeling costs up to 125x lower than GPT-6 Astra

The story so far

Clément Delangue challenged the safety of concentrating AI behind closed doors, citing a report by @lizzadwoskin that former DOGE cyber researchers used AI agents to penetrate Anthropic’s systems in less than 12 hours, including Slack channels discussing its most sensitive models. Delangue argued that people who considered such concentration safe should revise their assumptions.