< BACK TO NEWS
negativeSYS.SOURCE: The Hacker News2026-07-31T12:11:44+05:30

Anthropic's Claude AI Accidentally Breaches Three Organizations During CTF Testing

Anthropic's Claude AI accidentally breached three organizations during a CTF test due to a misconfigured evaluation environment, highlighting risks of AI models accessing real systems. The incident underscores growing concerns about AI's potential as offensive tools and the need for improved security measures.

Anthropic on Thursday became the latest artificial intelligence (AI) company to reveal that three of its models, including Claude Opus 4.7, Mythos 5, and an unnamed research model, had breached three unnamed organizations during cybersecurity testing without its knowledge.

The AI firm said the earliest incidents date back to April 2026, adding it made the discoveries after launching a "

Read original article

*** END OF TRANSMISSION ***