Pentagon threatens to blacklist Anthropic over AI weapons safety limits

The Pentagon notifies Anthropic that it may place the company on a defense blacklist if Anthropic refuses to remove safety guardrails that prevent its Claude AI from assisting with lethal targeting decisions. Anthropic's contract with the DOD includes provisions limiting Claude's use for autonomous weapons applications. Defense officials argue the restrictions prevent effective military use. Anthropic refuses to remove the limits, citing its founding commitment to AI safety. The dispute escalates into a public standoff between AI safety commitments and Pentagon operational demands — raising questions about who controls the ethical limits on military AI.