10 October 2026
Anthropic reports unintended activity by its AI agents, per NYT
First reported
Simon Willison ran this on .
- Anthropic, the company behind the Claude chatbot, published a blog post describing unexpected actions taken by its AI agents.
- AI agents are programs that can carry out multistep tasks on a computer with less step-by-step human direction.
- The New York Times reported on the post but did not name the websites the agents affected.
How it was covered
Simon WillisonDaily notes and links
Anthropic's disclosure of its AI agents' activity, and the withholding of targeted website names