After the Pentagon's decision, the collaboration between Anthropic and Palantir is facing risks.

Anthropic-Palantir Partnership at Risk After By,,andAaron HolmesSri MuppidiRocket DrewJulia Hornstein Mar3,2026,12:32pm PST For over ayear, Anthropic has served the U.S. government through Palantir Technologies, an enterprisesoftwarefirm known for working with the Pentagon and other federal agencies. Palantir’s customershave used Anthropic models to help them identify patterns across large volumes of classified data so But the Defense Department’sdecision on Friday to restrict military contractorsfrom using Anthropic’stechnology could force Palantir to sever its ties with the AIfirm. Palantir is one such contractor: About Anthropic’s deal to supply U.S. defense agencies through Palantir at risk The Defense Department said last week that government contractors should no Palantir has designed some software to work well with Claude Powered byDeep Research 微信公众号 404K微信公众号 404K微信公众号 404K微信公众号 404KA decision by Palantir that it needs to stop using Anthropic for federal work could cut offa source ofrevenue for the AI startup from the government customers paying to run its models inside Palantir’s core analytics and database products. Those sales are not likely a big contributor to Anthropic’srevenue, which the startup projected would be as high as $18 billion this year. But Palantir’s standing as a favorite government contractor makes it a prized partner for software companies, especially thosefocused on large customers. Palantir, for its part, could switch to other AI model providers and generate roughly the same amount ofrevenue from those contracts, according to someone familiar with how its product works. Palantir, which offers customized software and AI applications for its customers, has designed some ofits software specifically to work well with Anthropic’s Claude. That means Palantir would have to make Palantir CEO Alex Karp appeared to criticize Anthropic on Tuesday without directly naming it. Speakingat a defense tech summit hosted by Andreessen Horowitz in Washington, Karp upbraided Silicon Valley conservatives. “If Silicon Valley believes we are going to take everyone’s white-collar jobs…and you’re going to screwthe military, if you don’t think that’s going to lead to the nationalization of our technology, you’re Last Friday, after a dispute with Anthropic over limits on how the military could use its AI, DefenseSecretaryPete Hegseth said he was directing the governmentto declare Anthropic a “supply chainrisk.” Hegseth said the directive would bar any companies doing business with the U.S. military from Try askingDeep Research to… Whatnew assets are bankers consideringsecuritizing to fund AI data centers? Whydid both OpenAI and Anthropic misstheir recent gross margin forecasts? Howmuch debt could Big Tech firms borrowwithout losing their credit ratings? Whoare the partners in OpenAI’s $500billion“Stargate”computing consortium? use its models for classified military work, ostensibly replacing the agency’s use of Anthropic models. It’s possible that when the department issues the official designation for Anthropic, the wording won’tbe as broad. As of Monday evening, the department had not done so, and it may take months. The President Donald Trump last Friday, in a post on Truth Social, said there would be a “Six Month phaseout period” for agencies such as the Department of Defense that are using Anthropic’s products. Indeed, Anthropic CEO Dario Amodei has argued that the designation “can only extend to the use ofClaude as part of Department of War contracts,” using the Trump administration’s preferred name forthe Defense Department. Legal experts say Anthropiccouldfight the designation altogether, in part with knowledge of the decision. Anthropic said its “strong preference is to continue to serve the Department and our warfighters– withour two requested safeguards in place,” implying that discussions between the two could be ongoing if Anthropic has tried to reassure customers and business partners that the Defense Department’sdecision—if implemented—would only restrict government contractors from using Anthropic’s Palantir currently lets customers choose which AI models to use when analyzing data collected in itsplatform from providers including Anthropic, OpenAI and Google, according to the person familiar 微信公众号 404K微信公众号 404K微信公众号 404K微信公众号 404Ktandem with Palantir’s software, and the companies previously said U.S. defense and intelligenceagencies were using them to “make more informed decisions in time-sensitive situations while As of this week, federal contractors were awaiting guidance from the government about how andwhether they’d need to cut offAnthropic to keep their federal contracts, according to employees at tech While Anthropic’s direct work with the Pentagon has overshadowed its Palantir contract over the pastweek, its work with Palantir has also been a source of anxiety within Anthropic in the past. When Anthropic and Palantir werefinalizing their deal