Software stocks took another pounding overnight as AI lab Anthropic continued its warpath of high-velocity announcements which wipe billions of dollars off everyone else’s market caps.
The latest catalyst was Claude Mythos, a new frontier model the company unveiled this week through Project Glasswing, a cybersecurity initiative with the kind of quasi-mystical branding we’ve come to expect from the AI industry in general and Anthropic in particular.
Anthropic says Mythos is too dangerous to release to the general public — a lovely piece of marketing copy, if you can get it — but says it has proved particularly capable at cybersecurity tasks, having already uncovered thousands of previously unknown vulnerabilities across widely used software, including every major operating system and web browser.
Only around 40 organisations, including Microsoft, Google, Apple, Amazon and Nvidia — will get access through Glasswing. Today, Bloomberg reported that Treasury Secretary Scott Bessent and Federal Reserve chair Jerome Powell had summoned Wall Street bankers for an urgent meeting about a new era of “greater cyber risk” thanks to Mythos and similar models.
