By Stan Karanasios
Publication Date: 2026-04-13 08:02:00
These days, new, more powerful artificial intelligence (AI) models are announced fairly regularly: the latest version of ChatGPT or Claude or Gemini always has new features and new capabilities that their makers would like customers to try out.
But now Anthropic has announced a new model with great fanfare, but it only grants access to a select handful of users. In what The New York Times calls a “frightening warning sign” about the model’s power, the company has instead launched an initiative called “Project Glasswing” to use the model for good rather than evil.
Why? Early reports suggested that the model was able to move out of a closed testing “sandbox” with instructions and send an email to a researcher.
A little worrying perhaps. More importantly, Anthropic claims Mythos has uncovered software vulnerabilities and bugs “in every major operating system and every major web browser.”
Find hidden vulnerabilities
In one notable example…
