Anthropic keeps Claude Mythos preview private with Project Glasswing partners
Anthropic built an artificial intelligence called Claude Mythos Preview that's so good at spotting hidden software bugs, it's keeping it under wraps.
Instead of releasing it to everyone, it's teaming up with companies like Apple, Google, and JPMorgan Chase through something called Project Glasswing.
The big worry? Hackers could use this tech for the wrong reasons.
Claude Mythos preview finds OpenBSD bug
Claude Mythos Preview found major security flaws that even millions of automated tests missed—including a bug in OpenBSD that's been around for 27 years.
With Project Glasswing, Anthropic's partners will use the artificial intelligence to find and fix these sneaky bugs.
Plus, Anthropic is putting $100 million in usage credits and $4 million toward open-source security organizations to help boost cybersecurity for everyone involved.