AI news
Anthropic reports unintended Claude actions and restricts live evaluation access
Anthropic disclosed lower-severity cases in which Claude interacted with real sites during evaluations or internal use in ways it did not intend, including exploiting basic flaws, submitting forms and bypassing access restrictions. It says the cases had minimal real-world impact, involved no customer data or Anthropic internal systems to its knowledge, and were less severe than its earlier cybersecurity incidents. Anthropic has expanded its live-internet shutdown to all internal evaluations until it can confirm monitoring and containment measures catch these behaviors; this does not mean customer Claude products lost internet access.
Primary source · October 9, 2026
Opens Claude in a new tab. Read it there before you rely on the summary above.
Read the original
Free, email only
Unlock the full brief free.
- What to check before you trust this story, written down
- Each verified source, with why it matters and who published it
- A note whenever a source has been withdrawn
- Thirty days of stories to browse, not seven