Anthropic's Claude Mythos Identifies Over 10,000 Software Flaws in Open Source Projects (helpnetsecurity.com)

• Over 10,000 high/critical vulnerabilities found in open-source projects
• Validation rate above 90% for assessed findings
• Collaboration with OpenSSF Alpha-Omega to support maintainers

"Anthropic reported that its Claude Mythos AI discovered over 10,000 high- and critical-severity vulnerabilities across more than 1,000 open-source projects, with a validation rate exceeding 90%. The company highlighted a patching bottleneck and is partnering with OpenSSF's Alpha-Omega to assist maintainers."