Replying to
@GossiTheDog@cyberplace.social personally: nope
el reg recently talked about it, too
analyzed 1,061 publicly attributed AI-assisted vulnerability discoveries from Anthropic's Project Glasswing & the Berkeley Vulnerability Research Initiative, then cross-referenced them against its Known Exploited Vulnerability database.
The result: just 14, or 1.3 percent, have been confirmed as exploited in the wild, almost identical to the rate across all vulnerabilities in VulnCheck's dataset.
https://www.theregister.com/security/2026/07/28/ai-found-bugs-arent-proving-any-easier-to-exploit-despite-the-hype/5279637
