Trump blacklists Anthropic, OpenAI immediately steps in to sign Pentagon AI deal

The Trump administration officially places Anthropic on a defense procurement blacklist after the company refuses to remove AI safety restrictions on lethal targeting assistance. Within hours, OpenAI announces it has signed a new Pentagon contract to provide AI services — including for applications Anthropic declined. Sam Altman, who had previously called for AI regulation, now offers GPT models without the autonomous weapons guardrails Anthropic maintained. Critics note OpenAI's decision reverses its own earlier safety commitments. Dario Amodei says Anthropic's limits "are not negotiable."