Anthropic says its most powerful AI cyber model is too dangerous to release publicly — so it built Project Glasswing

Anthropic launched Project Glasswing, a $100 million AI cybersecurity initiative using its unreleased Claude Mythos Preview ...

CEO of billion-dollar company uses ‘taxi driver test’ to decide who gets hired - even for ‘perfect’ candidates

The CEO recounted rejecting a top hire whom the company had chased for months after they failed the ‘taxi driver test’.