OpenAI Signs Pentagon Deal While Trump Blacklists Anthropic
In the span of a single day, the Trump administration rewrote the rules of engagement between AI companies and the U.S. military — and nobody emerged more powerful than OpenAI.
On Friday, just hours after President Trump ordered federal agencies to immediately cease all use of Anthropic technology, OpenAI announced it had reached an agreement with the Pentagon to provide AI for classified systems. The timing was not coincidental. While Anthropic was being blacklisted, OpenAI was walking through the door the rival had been shut out of.
The dramatic reversal caps a week of escalating confrontation between the Defense Department and Anthropic, the AI startup behind Claude. The Pentagon had demanded that Anthropic loosen ethical guidelines on its AI systems — specifically, remove restrictions on mass domestic surveillance and autonomous weapons. When Anthropic refused, the administration struck back with extraordinary force.
The Blacklisting
Defense Secretary Pete Hegseth designated Anthropic as a supply-chain risk to national security — a designation normally reserved for foreign adversaries. No American company has ever been publicly classified this way.
The Government Services Administration followed suit, terminating its contracts with Anthropic. The company's existing $200 million, two-year agreement with the Pentagon will be allowed to continue for a maximum of six months during a transition period.
Anthropic fired back late Friday: We will challenge any supply chain risk designation in court, calling the designation an unprecedented action.
The OpenAI Pivot
Within hours of Anthropic's blacklisting, OpenAI CEO Sam Altman announced the company's new Pentagon partnership.
Two of our most important safety principles are prohibitions on domestic mass surveillance and human responsibility for the use of force, including for autonomous weapon systems, Altman wrote. The Pentagon agrees with these principles and we put them into our agreement.
The announcement capped a remarkable 48 hours for OpenAI. Earlier on Friday, the company revealed it had raised $110 billion in private funding — one of the largest private funding rounds in history. Amazon contributed $50 billion, while Nvidia and SoftBank each invested $30 billion, valuing OpenAI at $730 billion.
The Split
The Pentagon wanted Claude AI capabilities for mass domestic surveillance and fully autonomous weapons systems. Anthropic refused. OpenAI, by contrast, appears to have negotiated terms that satisfied both the Pentagon's operational needs and its own stated safety principles.
What's at Stake
The confrontation is a test case for whether AI companies can maintain ethical boundaries when confronted with the full weight of the federal government.
Anthropic staked its reputation on being the most safety-forward of the major AI labs. That positioning helped differentiate it but now faces its ultimate test.
The company has vowed to fight the designation in court. If it prevails, it could establish important precedents for AI company autonomy. If it loses, it may find itself locked out of any company that does business with the government.
Meanwhile, OpenAI has doubled down on its relationship with the defense establishment. The company now has a Pentagon deal, $110 billion in fresh capital, and the implicit endorsement of an administration that just destroyed its primary competitor in the space.