Overview
- Anthropic, which announced Project Glasswing on Tuesday, opened Claude Mythos Preview to a small group of tech and security firms for defensive work.
- The company says the model found thousands of serious flaws across major operating systems and browsers, including a 27-year OpenBSD bug and a 16-year FFmpeg issue.
- Access remains restricted as Anthropic withholds a broad release and commits up to $100 million in usage credits plus $4 million for open-source security groups.
- Researchers report the system can spot bugs and write working exploits with little guidance, which they say could help defenders but also increases the risk of automated attacks.
- Anthropic says it briefed U.S. agencies on the tool’s capabilities as it contests a federal supply‑chain risk designation that a judge recently paused.