Anthropic's Mythos model sparks alarm

The artificial intelligence community is on high alert after Anthropic's experts warned that Mythos could potentially hack the systems underlying modern computing.

Under a secretive initiative dubbed Project Glasswing, over 50 selected organizations are now testing the model's capabilities in a controlled environment.

The breakthrough has prompted calls for greater international dialogue on AI safety, with Nvidia CEO Jensen Huang urging cooperation between the U.S. and China.

While the model offers unprecedented defensive potential, its ability to automate complex cyberattacks has forced a rethink of digital security protocols.