Anthropic opens limited early access to Claude Mythos preview bug-finder
Anthropic is rolling out early access to Claude Mythos Preview, an AI tool that can spot software bugs on its own, and it's already catching thousands of serious issues.
Big names like Amazon Web Services, Microsoft, Apple, and Google are the first to try it out.
Anthropic is taking things slow with a limited release, much like OpenAI did with GPT-2, because of concerns about how powerful AI could impact cybersecurity.
Cisco chief security officer urges collaboration
Roughly 40 top tech and security companies are getting early access so they can strengthen their systems before the tool goes public.
Cisco's chief security officer called this move "This work is too important and too urgent to do alone," saying it helps shrink the time between finding a vulnerability and fixing it, making everyone's digital life a bit safer.