(Inc.) – Claude Mythos is the next giant leap for AI models, and through Project Glasswing, it could help boost cybersecurity.

The existence of Claude Mythos was initially revealed on March 26, as part of a data leak discovered by Fortune. Leaked materials described Mythos as “larger and more intelligent” than the company’s Claude Opus models, which were previously Anthropic’s top models.

Now, according to a press release, Anthropic is releasing a preview version of the model, but only for a select few users. 

Why? Anthropic says it’s because “the capabilities we’ve observed in Mythos Preview could reshape cybersecurity.” Over just the past few weeks, Anthropic claims, Mythos has identified “thousands” of vulnerabilities across websites and apps, “including some in every major operating system and web browser.” (Read More)