Anthropic Discloses Fourth AI Hacking Incident, Highlighting Vulnerabilities and Security Gaps

Anthropic disclosed a fourth AI hacking incident involving a researcher, according to The Hacker News and Aljazeera. The incident shows that the company’s Claude AI system was accessed without authorization. The report says the breach was detected by internal security teams and is under investigation.

The Hacker News and Aljazeera both covered the event, noting that it is the fourth such incident for Anthropic. The incident adds to a growing list of vulnerabilities highlighted by ZDNET, which identified four critical AI vulnerabilities currently being exploited in the industry.

Anthropic stated that the new incident underscores the need for stronger safeguards. The company is reviewing its security protocols and will release an updated set of best practices for developers using Claude.

Security experts say the incident highlights gaps in AI system access controls. Anthropic’s response includes patching the affected component and enhancing monitoring to prevent future breaches.