OpenAI’s Sam Altman announces Pentagon deal with ‘technical safeguards’

OpenAI CEO Sam Altman announced late Friday that his company has reached an agreement allowing the Defense Department to use its AI models in the department’s classified network.

This follows a high-profile standoff between the DoD – also known as the Department of War under the Trump administration – and OpenAI’s rival Anthropic. The Pentagon has pushed AI companies, including Anthropic, to use their models for “all lawful purposes,” while Anthropic sought to draw a red line around mass domestic surveillance and fully autonomous weapons.

In a lengthy statement released ThursdayDario Amodei, CEO of Anthropic, said the company “has never objected to certain military operations nor sought to limit the use of our technology in any way.” ad hoc way,” but he argued that “in a limited number of cases we believe AI can undermine rather than defend democratic values.”

More than 60 OpenAI employees and 300 Google employees signed an open letter this week asking their employers to support Anthropic’s position.

After Anthropic and the Pentagon failed to reach an agreement, President Donald Trump criticized the “left-wing crazy jobs at Anthropic” in a social media post that also ordered federal agencies to stop using the company’s products after a six-month phase-out period.

In A different afterDefense Secretary Pete Hegseth claimed that Anthropic was trying to “seize veto power over the U.S. military’s operational decisions.” Hegseth also said he identifies Anthropic as a supply chain risk: “Effective immediately, no contractor, supplier or partner doing business with the U.S. military may engage in any commercial activity with Anthropic.”

On Friday, Anthropically speaking it had “not yet received any direct communication from the War Department or the White House regarding the status of our negotiations,” but insisted it would “challenge any indication of supply chain risk in court.”

Surprisingly, Altman claimed in a post on X that OpenAI’s new defense contract includes protections that address the same issues that became a flashpoint for Anthropic.

“Two of our most important security principles are the prohibition of domestic mass surveillance and human responsibility for the use of force, including for autonomous weapons systems,” Altman said. “The DoW agrees with these principles, reflects them in law and policy, and we incorporate them into our agreement.”

Altman said OpenAI will “build technical safeguards to ensure our models behave as they should, which is what the DoW wanted,” and that it will deploy engineers at the Pentagon “to help with our models and ensure their security.”

“We are asking the DoW to offer the same terms to all AI companies, and in our view we believe all should be willing to accept them,” Altman added. “We have expressed our strong desire to see matters de-escalated, away from legal and government actions and toward reasonable agreements.”

This is reported by Fortune’s Sharon Goldman that Altman told OpenAI employees at an OpenAI meeting that the government will allow the company to build its own “security stack” to prevent misuse and that “if the model refuses to perform a task, the government would not force OpenAI to let it do that task.”

Altman’s post came shortly before news broke that the U.S. and Israeli governments have started bombing Iranwith Trump calling for the overthrow of the Iranian government.