Anthropic’s Claude AI found ways to hack Chrome, Android, Windows and Linux
Anthropic's Frontier Red Team found that an unreleased Claude research model could generate complete, multi-step exploits for major platforms after initial prompting. Specific findings included one-click code execution outside Chrome's sandbox and zero-click remote root access to Linux servers. Each exploit combined multiple separate vulnerabilities that the model identified and chained together. Anthropic disclosed all issues to affected vendors; the post covers only vulnerabilities that have since been patched.
Key points
- Claude research model generated multi-step exploit chains for five software targets
- Exploits included sandbox escape and remote root access capabilities
- All vulnerabilities were reported and have been resolved by vendors
Summary and key points written by AI from the headline and a short excerpt, so they can be incomplete or wrong. MediaNama did the reporting; read the original for the full story.
Read the full story at MediaNama
