Anthropic, for the first time, has officially spoken on its showdown with the Pentagon, following a high-stakes “ultimatum” from the US Department of War. In a statement, the company described its recent meeting with Defense Secretary Pete Hegseth as a “good-faith conversation” focused on balancing military needs with AI safety. The statement comes after the AI company has been given a Friday (February 27) at 5:00 pm deadline to scrap its internal safety rules or face severe government penalties.An Anthropic spokesperson confirmed that CEO Dario Amodei met with Secretary Hegseth at the Pentagon to discuss the future of their partnership, Axios reported“During the conversation, Dario expressed appreciation for the Department’s work and thanked the Secretary for his service,” an Anthropic spokesperson said, adding, “We continued good-faith conversations about our usage policy to ensure Anthropic can continue to support the government’s national security mission in line with what our models can reliably and responsibly do.”“The only reason we’re still talking to these people is we need them and we need them now. The problem for these guys is they are that good,” a Defense official told Axios ahead of the meeting.According to Axios, Anthropic said it is willing to adapt its usage policies for the Pentagon but will not allow its model for the mass surveillance of US citizen or the development of weapons that fire without human involvement.

Why the Pentagon has given Anthropic a deadline

Previous reports have said that the Pentagon and Anthropic have an issue over the use of AI company’s solutions in US military for “all lawful purposes”. These include mass surveillance of American citizens and fully autonomous weapons. In a tense meeting, Hegseth delivered a strict ultimatum to Anthropic CEO Dario Amodei: drop the company’s strict AI safety rules by Friday, or face severe government retaliation. Citing sources, news agency Reuters reported that Anthropic has no intention of backing down on its usage restrictions as the AI startup has refused to remove safeguards.