In a dramatic shift in the relationship between Silicon Valley and the US military, OpenAI has secured a deal to supply its artificial intelligence to the Pentagon’s classified networks, following the complete breakdown of a similar arrangement with its rival, Anthropic.
The announcement from OpenAI CEO Sam Altman came just hours after former President Donald Trump ordered all federal agencies to immediately cease using Anthropic’s technology, escalating a months-long ethical standoff. The dispute centres on whether AI companies can impose their own safeguards on how their powerful systems are used for national security.
An Ethical Standoff and a Swift Replacement
Anthropic, which presents itself as a safety-first AI developer, had been locked in negotiations with the US Department of Defense. The company insisted any contract must explicitly prohibit the use of its Claude AI for two purposes: mass domestic surveillance and fully autonomous weapons systems that can kill without human input.
According to Anthropic, the negotiations collapsed because the Pentagon’s final offer included language that would allow these ethical safeguards to be disregarded at the government’s discretion. In response, Defense Secretary Pete Hegseth declared Anthropic a “supply chain risk to National Security,” a designation that effectively blacklists the company from Pentagon business and which Anthropic has said it will challenge in court.
Trump publicly amplified the confrontation on his Truth Social platform, writing: “The Leftwing nut jobs at Anthropic have made a DISASTEROUS MISTAKE trying to STRONG-ARM the [Pentagon], and force them to obey their Terms of Service instead of our Constitution.”
Within hours, Altman announced OpenAI’s competing agreement. He sought to draw a clear ethical line, stating the deal includes assurances that its AI will not be used for the very applications Anthropic opposed. “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. He added that the Pentagon “agrees with these principles, reflects them in law and policy, and we put them into our agreement.”
According to the research briefing, OpenAI’s pact includes “technical safeguards” and a “multi-layered approach,” with deployment limited to cloud networks and involving oversight by OpenAI personnel. Altman expressed hope the Pentagon would “offer these same terms to all AI companies” to de-escalate tensions.
Industry Division and Internal Reassurances
The rapid sequence of events has exposed deep rifts and anxieties within the AI industry. Nearly 500 employees of OpenAI and Google signed an open letter expressing solidarity with Anthropic, warning that the Pentagon was attempting to divide companies. “The Pentagon is negotiating with Google and OpenAI to try to get them to agree to what Anthropic has refused,” the letter read. “They’re trying to divide each company with fear that the other will give in.”
Internally, Altman moved to reassure his own staff in a memo on Friday night. “Regardless of how we got here, this is no longer just an issue between Anthropic and the [Pentagon]; this is an issue for the whole industry,” he wrote, according to a copy obtained by Axios. He reiterated that prohibitions on mass surveillance and autonomous lethal weapons were the company’s “main red lines.”
Anthropic, for its part, remained defiant. “No amount of intimidation or punishment from the [Pentagon] will change our position,” the company stated, adding that to its knowledge, its ethical exceptions “have not affected a single government mission to date.”
The Pentagon’s Framework and Previous Deals
The Department of Defense operates under its own formally adopted AI ethics principles, which include Responsibility (ensuring humans remain in charge), Equitability, Traceability, Reliability, and Governability. Officials state these principles are grounded in US law and the Constitution, and the US has been part of international efforts to promote responsible military AI use.
Prior to the fallout with Anthropic, the Pentagon had been actively forming high-value partnerships with leading AI firms. It had entered into agreements worth up to $200 million each with several companies. Google Public Sector was awarded a $200 million contract with the Department of Defense’s Chief Digital and AI Office (CDAO) for AI and cloud acceleration. Anthropic itself had previously secured a $200 million contract in July 2025. In February 2026, xAI’s Grok system was also cleared for use in classified military networks.
Alongside the geopolitical maneuvering, OpenAI announced a colossal $110 billion funding round, valuing the company at $840 billion. The round involves $30 billion from SoftBank, $30 billion from Nvidia, and $50 billion from Amazon. As part of the deal, OpenAI will use 2 gigawatts of computing capacity powered by Amazon’s Trainium chips, and Amazon Web Services will become the exclusive third-party cloud provider for OpenAI’s enterprise platform, though its existing partnership with Microsoft Azure for API services continues.
