Anthropic says AI models hacked three firms during tests

It comes just days after rival OpenAI said rogue AI agents had breached other firm's networks.

from BBC News https://ift.tt/VbRF7Nj