Claude Finds Zero-Day Vulnerabilities in Ghost CMS and the Linux Kernel in Live Demo
Anthropic researcher Nicholas Carlini demonstrated Claude discovering previously unknown vulnerabilities in production software during a conference talk — raising urgent questions about AI's dual-use potential in cybersecurity.
In what may be the most consequential live demo in recent AI history, an Anthropic researcher demonstrated Claude independently discovering zero-day vulnerabilities in both the Ghost publishing platform and the Linux kernel. As @_xjdr reported from the event, "someone at ANTHROPIC just showed CLAUDE finding ZERO DAY vulnerabilities... claude has found zero day in Ghost... then did the exact, same thing to the linux kernel."
Unlock the full briefing
Get every story in today's briefing, the full archive, and the daily AI intelligence brief.
All stories today
Full archive
Daily brief
Cancel anytime. Payments powered by Stripe.