'After one month, most partners have each found hundreds of critical- or high-severity vulnerabilities': Anthropic claims Mythos has found over ten thousand major security vulnerabilities across 'the most systemically important software in the world'
- Anthropic reported Mythos Preview uncovered over 10,000 high‑ and critical‑severity vulnerabilities in under two months, with Cloudflare alone finding 2,000 bugs
- Independent validation confirmed 90% of assessed findings as real, though critics argue the breakthrough may stem from massive compute and workflows rather than unique reasoning
- The bottleneck has shifted from discovery to verification, disclosure, and patching, as AI now surfaces vulnerabilities faster than organizations can remediate them
In less than two months, Anthropic and friends have apparently discovered more than ten thousand critical and high-level security vulnerabilities using the famed Mythos Preview artificial intelligence tool.
In a brief update on the state of the project, published late last week, Anthropic said that since the release of the tool, roughly 50 organizations that got to use it each found “hundreds” of vulnerabilities.
“Several have told us that their rate of bug-finding has increased by more than a factor of ten,”...
Copyright of this story solely belongs to techradar.com. To see the full text click HERE
