What happened
Anthropic unveiled Claude Mythos Preview, a new general-purpose language model. This model autonomously identifies and exploits zero-day vulnerabilities across every major operating system and web browser, including flaws undetected for decades. In response, Anthropic launched Project Glasswing, an initiative to use Mythos Preview for defensive security. Anthropic committed up to $100M in usage credits for participants.
Why it matters
This development shifts the cybersecurity landscape, granting a single entity exclusive, albeit temporary, power to exploit widespread software. Security architects and platform engineers face immediate pressure to proactively test and patch systems using advanced AI tools. Anthropic's commitment of $100M in usage credits for defenders highlights the significant investment required for this capability, impacting procurement teams' budgets for vulnerability detection. This follows previous concerns where Anthropic denied AI sabotage capability, now demonstrating the dual-use nature of advanced AI.
Subscribe for Weekly Updates
Stay ahead with our weekly AI and tech briefings, delivered every Tuesday.
