OpenAI signs deal with Pentagon after Trump blacklists Anthropic

OpenAI CEO Sam Altman said late Friday that his company had agreed to terms with the Defense Department regarding the use of its artificial intelligence models, shortly after President Donald Trump said the government would not work with AI rival Anthropic.

“Tonight, we reached an agreement with the Department of War to deploy our models on their classified network,” Altman wrote in a post on X. “In all our interactions, the DoW has demonstrated a deep respect for safety and a desire to partner to achieve the best possible outcome.”

Altman’s post comes at the end of a dramatic week for the AI ​​industry, which has been at the center of a political debate surrounding how its models can be used. Earlier in the day, Defense Secretary Pete Hegseth designated Anthropic as a “supply-chain threat to national security” after weeks of tense negotiations. The label is typically reserved for foreign adversaries, and the DoD requires vendors and contractors to certify that it does not use anthropogenic models.

President Trump directed every federal agency in the US to “immediately stop” using Anthropic’s technology.

Anthropic is the first lab to deploy its samples across the DOD’s classified network and is trying to negotiate the ongoing terms of its contract with the agency before talks collapse. The company wanted assurances that its models would not be used for fully autonomous weapons or mass surveillance of Americans, but the DoD wanted Anthropic to agree to allow the military to use the models in all legitimate use cases.

OpenAI shared the same “red lines” as Anthropic, Altman told employees in a memo Thursday. The DOD has agreed to its sanctions, he said in his post on Friday.

“Two of our core security principles are domestic mass surveillance and prohibitions on human responsibility for the use of force, including autonomous weapons systems,” Altman wrote. DoW agrees with these principles, reflects them in law and policy and we put them into our contract.

It wasn’t immediately clear why the DoD agreed to accommodate Open AI, not Anthropic, though government officials have criticized Anthropic for months for being overly concerned about AI safety.

Altman said OpenAI builds “technical safeguards to make sure its models behave the way they’re supposed to,” and the company deploys staff to “help our models and make sure they’re safe.”

“We are asking the DoW to issue similar rules to all AI companies, which we think everyone should be willing to accept,” Altman wrote. “We have expressed our strong desire to see matters escalate away from legal and government action and towards reasonable agreements.”

Anthropic said in a statement Friday that it was “deeply saddened” by the Pentagon’s decision to label the company a supply chain risk. It said it intends to challenge that designation in court.

Watch: Hegseth directs Pentagon to report anthropic supply-chain risk