German cyber agency braces for 'significant disruption' from Anthropic's AI hacking tech

A powerful new AI model unveiled by Anthropic is a "paradigm change in the nature of cyber threats," the head of Germany's cyber agency told POLITICO.

Claudia Plattner, president of the German Federal Office for Information Security (BSI), said her agency is in "active dialogue" with Anthropic about the company's new model, called Mythos, that is believed to be outperforming humans at finding and exploiting software bugs and glitches.

German agency officials take the U.S. company's new model "very seriously" and "anticipate significant disruption," Plattner said.