OpenAI Strikes Pentagon Deal Hours After Trump Blacklists Anthropic

When Anthropic CEO Dario Amodei refused to budge on his company’s ethical red lines, he knew the consequences could be severe. What he didn’t expect was how quickly the landscape would shift around him. Within hours of the Trump administration blacklisting Anthropic for its refusal to allow unrestricted military use of its AI, rival OpenAI announced it had struck its own deal with the Pentagon—complete with the very same safeguards Anthropic had fought to maintain.

“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.” — Sam Altman, OpenAI CEO

The Deadline That Changed Everything

The standoff reached its climax on Friday afternoon when the Pentagon’s 5:01 p.m. ET deadline came and went. Defense Secretary Pete Hegseth had demanded that Anthropic remove restrictions preventing its Claude AI from being used for mass domestic surveillance or fully autonomous weapons. Amodei refused, calling such uses “outside the bounds of what today’s technology can safely and reliably do.”

The administration’s response was swift and severe. President Trump directed every federal agency to immediately cease using Anthropic technology, declaring on Truth Social: “We don’t need it, we don’t want it, and will not do business with them again!” The Pentagon designated Anthropic a “supply chain risk to national security”—a label typically reserved for foreign adversaries like Huawei.

The six-month transition period creates immediate operational challenges. Claude had been integrated into classified military systems supporting sensitive intelligence work, weapons development, and operational planning. Defense officials now face the complex task of extracting Anthropic’s technology from these critical workflows without degrading capabilities.

OpenAI’s Strategic Pivot

While Anthropic absorbed the administration’s blows, OpenAI moved decisively. CEO Sam Altman announced that his company had reached an agreement for the Pentagon to use OpenAI models in classified networks—with explicit safeguards mirroring Anthropic’s position.

The technical safeguards embedded in OpenAI’s contract specifically prohibit domestic mass surveillance and require human oversight for any use of force, including autonomous weapons. Altman emphasized that these weren’t concessions but core principles: “Humans should remain in the loop for high-stakes automated decisions.”

The timing raised eyebrows across Silicon Valley. Altman had spent the morning expressing solidarity with Anthropic, telling CNBC he didn’t “personally think the Pentagon should be threatening DPA against these companies.” By evening, OpenAI had secured the classified access that Anthropic had just lost.

“For all the differences I have with Anthropic, I mostly trust them as a company, and I think they really do care about safety.” — Sam Altman on CNBC, hours before announcing OpenAI’s Pentagon deal

Industry Solidarity vs. Competitive Reality

The conflict has produced unusual alliances. More than 300 Google employees and over 60 OpenAI staffers signed an open letter urging their leaders to stand with Anthropic. The letter, titled “We Will Not Be Divided,” warned that the Pentagon was “trying to divide each company with fear that the other will give in.”

Google DeepMind Chief Scientist Jeff Dean added his voice, posting on X that “mass surveillance violates the Fourth Amendment and has a chilling effect on freedom of expression.” While Dean spoke as an individual, his statement carried weight given Google’s own Pentagon negotiations.

Yet competitive dynamics proved stronger than solidarity. With Anthropic sidelined, the Pentagon has accelerated talks with Google and OpenAI about expanding their models into classified environments. Elon Musk’s xAI, which already gained approval for classified military use of its Grok model this week, stands ready to capture market share.

The Fundamental Question

Beyond the immediate corporate maneuvering, the dispute raises profound questions about AI governance. Can private companies set ethical boundaries on how their technology is used by government? Or does national security require that the military have “full, unrestricted access” to AI capabilities, as Secretary Hegseth demanded?

Anthropic’s legal challenge will test these boundaries. The company has vowed to fight the supply chain risk designation in court, arguing it “would both be legally unsound and set a dangerous precedent for any American company that negotiates with the government.”

Retired Air Force General Jack Shanahan, former leader of the Pentagon’s AI initiatives, offered a measured assessment: “Painting a bullseye on Anthropic garners spicy headlines, but everyone loses in the end.” Shanahan noted that Claude’s existing restrictions were “reasonable” and that current AI models are “not ready for prime time in national security settings,” particularly for autonomous weapons.

The coming months will reveal whether Anthropic’s principled stand costs it the government market permanently, or whether the administration’s hardline approach softens. For now, one thing is clear: the battle over who controls AI’s most powerful capabilities has moved from conference rooms to the front pages—and the stakes extend far beyond any single company’s bottom line.

This article was reported by the ArtificialDaily editorial team. For more information, visit TechCrunch and NPR.