Anthropic works with White House after AI security scare

Weeks after forcing one of its top AI models offline, Washington wants to write the rules with the company it just punished.