TechThe Register2h ago
Claude Opus wrote a Chrome exploit for $2,283
TL;DRAI models can find Chrome security holes; Anthropic restricted its bug-finding model over safety concerns.
Why it matters: Public AI systems already pose real exploit risks, validating debates over AI capabilities access and responsibility.
Pause your Mythos panic because mainstream models anyone can use already pick holes in popular software Anthropic withheld its Mythos bug-finding model from public release due to concerns that it would enable attackers to find and exploit vulnerabilities before anyone could…
Read full articleSource: The Register · Opens in new tab