Quick Summary
- OpenAI pursuing partnership to implement AI systems on NATO’s non-classified infrastructure following CEO Sam Altman’s clarification about access levels
- The company recently finalized an agreement to integrate AI capabilities into the Pentagon’s secure classified infrastructure
- Trump administration directed federal departments to terminate Anthropic partnerships, eliminating contracts valued above $200 million
- Anthropic’s removal followed its refusal to grant the Pentagon unrestricted AI model access
- Federal departments including State, Treasury, and HHS are transitioning from Anthropic’s services to OpenAI’s platform
OpenAI is pursuing a collaboration with NATO while standing behind its recently announced Pentagon partnership, as the federal government phases out competitors like Anthropic from government agreements.
During an internal company gathering Tuesday, OpenAI CEO Sam Altman defended the Pentagon partnership to employees, though he acknowledged the public announcement appeared hastily executed and potentially opportunistic.
“We were genuinely trying to de-escalate things and avoid a much worse outcome, but I think it just looked opportunistic and sloppy,” Altman said in a memo posted on X.
The previous week, OpenAI finalized a landmark agreement with the U.S. Department of Defense to integrate its artificial intelligence systems into secure government infrastructure. This arrangement materialized shortly after President Trump directed federal departments to discontinue Anthropic’s AI solutions.
The administration terminated over $200 million worth of Anthropic agreements. Defense Secretary Pete Hegseth characterized Anthropic as representing a national security “supply chain risk.”
The Reasons Behind Anthropic’s Federal Contract Termination
Anthropic CEO Dario Amodei explained his organization lost federal partnerships after declining to remove safeguards limiting military applications of its artificial intelligence. The company resisted Pentagon requirements for unrestricted model deployment.
Anthropic had specifically maintained restrictions against utilizing its AI for widespread domestic surveillance operations or fully autonomous weapon systems. While the Pentagon stated no interest in those particular applications, it insisted on permitting all legally authorized AI implementations.
OpenAI’s revised Pentagon partnership specifies its AI “shall not be intentionally used for domestic surveillance of U.S. persons and nationals.” The NSA additionally verified that intelligence organizations would not utilize AI services under this arrangement.
Altman expressed no regrets about partnering with the Defense Department, though he wished the timing of the public announcement had been better considered. He acknowledged to employees the announcement appeared to capitalize on Anthropic’s departure.
“To try so hard to do the right thing and get so absolutely personally crushed for it is really painful,” Altman said at the staff meeting.
NATO Partnership Negotiations
Concurrently, OpenAI has entered negotiations to implement its technology across NATO’s non-classified network infrastructure. NATO represents a 32-nation military coalition.
Altman initially informed employees that OpenAI was exploring deployment across all NATO classified networks. A company representative subsequently issued a correction, clarifying the contract opportunity involves only unclassified network systems.
NATO officials have not provided comment on the discussions.
Federal departments including the State Department, Treasury, and the Department of Health and Human Services are likewise discontinuing Anthropic’s AI services pursuant to the new executive directive.
Anthropic counts Amazon and Google among its major investors. OpenAI’s backers include Microsoft and Amazon, along with other prominent investors.
OpenAI has publicly stated its position that Anthropic should not receive a “supply chain risk” designation from federal authorities.
Altman noted the government was prepared to grant OpenAI significant influence regarding technology deployment. “We have built a technology that is going to be the fundamentally most important tool for the government and governments around the world,” he said.


