Proper in the midst of the continuing feud between the Silicon Valley AI firm Anthropic and the U.S. Division of Protection over whether or not the army will use—or not use—Anthropic’s massive language fashions is one more firm: Palantir.
Palantir, the Miami-based information analytics and synthetic intelligence platform, is a key software program supplier for the Division of Protection—and the principle channel by which the Division has been utilizing Anthropic’s massive language mannequin, Claude.
“We are legitimately still in the middle of all this,” CEO Alex Karp mentioned in an interview with Fortune on the sidelines of the corporate’s twice-a-year AIP convention on Thursday. “It’s our stack that runs the LLMs.”
Karp says he had been in quite a few discussions with all events concerned—discussions he declined to provide specifics about, as he says he doesn’t wish to “out conversations” or “bash people.”
However Karp does wish to make one factor clear: The Protection Division shouldn’t be utilizing AI for home mass surveillance on U.S. residents—and, to his data, it has no plans to.
“Without commenting on internal dialogs, there was never a sense that these products would be used domestically,” Karp mentioned. “The Department of War is not planning to use these products domestically. That’s a completely different kettle of fish… The terms the Department of War wants are completely focused on non-American citizens in a war context.”
Palantir has an unlimited enterprise doing work for the U.S. authorities, together with the DoD. Anthropic partnered with Palantir in 2024 to supply its AI know-how to the DoD through Palantir. Anthropic additionally started working instantly with the DoD final yr to create a model of its know-how designed for the Protection Division.
The contentious back-and-forth between Anthropic and the Protection Division has been ongoing since round January, and the 2 sides don’t agree on what set it off. Statements that Undersecretary of Protection for Analysis and Engineering Emil Michael made final week allege that Palantir had notified the Pentagon that Anthropic was inquiring about whether or not its fashions had been used for the U.S. army mission to seize Venezuelan President Nicolás Maduro. (Anthropic has refuted this characterization, asserting it hasn’t mentioned the usage of Claude for particular operations “with any industry partners, including Palantir, outside of routine discussions on strictly technical matters”). Ever since, the 2 sides have been locked in a battle over whether or not Anthropic can write contractual limits on how its fashions are used.
Anthropic CEO Dario Amodei has revealed a number of weblog posts on the matter, together with an preliminary assertion on the finish of February asserting that the Protection Division had refused to simply accept safeguards that its LLMs not be used for home mass surveillance or the deployment of totally autonomous weapons. Pete Hegseth, the Secretary of Protection, later designated Anthropic a “supply-chain risk,” threatening most of the firm’s business relationships, and prompting Anthropic to sue the Pentagon over the designation.
‘Totally in favor’ of home phrases of engagement
Civil liberties teams, nonetheless, proceed to accuse the corporate of doing the alternative—by serving to the federal government surveil. The corporate’s relationship with U.S. Immigration and Customs Enforcement, specifically, which started underneath the Obama Administration, has invited intense scrutiny and criticism from each exterior critics and the corporate’s personal staff—criticism that has solely escalated over the past yr because the Trump Administration has pushed ICE into an aggressive crackdown in cities like Minneapolis.
Karp instructed Fortune he’s “very sympathetic with arguments against using these products inside the U.S.” and mentioned that he’s “totally in favor” of setting phrases of engagement and limits to how home companies can use synthetic intelligence.
“Quite frankly, I think we should self-impose them,” Karp mentioned of those phrases of engagement. “The Valley should have a consortium: This is what we’re going to do, and this is what we’re not going to do,” he mentioned.
However Karp drew a pointy distinction between whether or not tech corporations ought to set phrases with home companies and whether or not they need to set them with the Division of Protection, which is primarily targeted on managing the US’ relationships with different nations and its adversaries.
“What we’re talking about now is using products vis-a-vis someone who’s trying to kill our service members,” Karp mentioned, noting that he personally helps “wide license” of utilization for the Division of Protection particularly.
“If we knew China and Russia and Iran wouldn’t build them, I would be in favor of very heavy—very heavy—legal constraints,” Karp mentioned. However he factors out that American adversaries will construct them and use them in opposition to the U.S. anyway. “I don’t think this is an opinion. I think this is a fact, and that fact means I think the Department of War should have wide license to use these products.”
