Anthropic’s Mythos model found vulnerabilities in classified US government systems, official says

Anthropic has raised concerns over how the U.S. military would use its AI, while the administration has restricted the use of some of its models