Jump to content

Anthropic–United States Department of Defense dispute

From Deep Politics Wiki


Full article adapted from Wikipedia revision 1375722324 by Wikipedia contributors. Text is available under CC BY-SA 4.0. Formatting has been adapted; article text and references are retained. Snapshot imported 2026-09-25.

Pete Hegseth
Dario Amodei
Hegseth, the United States secretary of defense, has publicly rebuked Anthropic chief executive Dario Amodei's approach to artificial intelligence.

Since January 2026, the United States Department of Defense has conflicted with the artificial intelligence company Anthropic over the use of its products for military purposes and mass domestic surveillance.

Background

Artificial intelligence in the U.S. military

Main article: Use of artificial intelligence by the United States Department of Defense
See also: Artificial intelligence arms race

The United States Department of Defense began developing lethal autonomous weapons as early as the Reagan administration.[1] The Department of Defense established a policy on the use of artificial intelligence in 2012, Directive 3000.09.[2] Efforts to utilize artificial intelligence intensified under the term of Secretary Ash Carter.[3] The Department of Defense's use of artificial intelligence for Project Maven prompted concerns within Google in 2018, leading to protests and mass resignations.[4]

Anthropic in the second Trump administration

In Donald Trump's second presidency, Anthropic publicly disagreed with the administration's policies and initiatives. In January 2025, Anthropic CEO Dario Amodei criticized the artificial intelligence investment project Stargate as "chaotic" and opposed Trump's rescission of president Joe Biden's Executive Order on Artificial Intelligence, but noted that Anthropic had held discussions with Trump officials about artificial intelligence policy.[5] Amid discussions over the One Big Beautiful Bill Act, Anthropic privately lobbied for Congress to vote against a bill preventing states from regulating artificial intelligence and expressed opposition to an artificial intelligence agreement signed among Gulf states in Trump's visit to the Middle East in May. According to Semafor, Trump officials chastised Anthropic's hiring of several officials involved in the Biden administration, including Elizabeth Kelly, the former director of the Artificial Intelligence Safety Institute; Tarun Chhabra, the coordinator for technology and national security in the National Security Council; and Ben Buchanan, Biden's advisor for artificial intelligence.[6] The following month, Amodei wrote an op-ed in The New York Times describing the artificial intelligence regulation bill, then tied to the One Big Beautiful Bill Act, as "far too blunt an instrument".[7]

Prior to the dispute, the Trump administration had integrated Anthropic's services. By November 2024, Anthropic had already partnered with Palantir and Amazon Web Services, companies that offered services with FedRAMP authorization. In the Biden administration, Anthropic had reached an agreement with the AI Safety Institute and had participated in a nuclear information safety evaluation.[8] The Department of Homeland Security authorized its workers to use commercial artificial intelligence systems, including Anthropic's Claude, until May 2025.[9] Through its interoperability with Palantir, a company heavily involved in data analysis and analytics at the Department of Defense, Anthropic's technology achieved relatively widespread usage in the U.S. military.[10] The following month, Anthropic announced that it would allow national security customers to use Claude Gov.[11][12] Anthropic's orthogonal usage policy to the surveillance systems implemented at the Federal Bureau of Investigation, the Secret Service, and Immigration and Customs Enforcement led to a conflict between Anthropic and the Trump administration by September.[13] That month, Amodei criticized Trump's approach to export restrictions on semiconductors.[14]

Anthropic's strategy has mirrored Amodei's views toward Trump; in a Facebook post ahead of the 2024 presidential election, Amodei urged his associates to vote for vice president Kamala Harris over Trump, describing him as a "feudal warlord". As the Trump administration targeted law firms, Amodei cut ties with the firms Skadden, Arps, Slate, Meagher & Flom and Latham & Watkins, which reached agreements with the Trump administration to avoid punishment. David Sacks, Trump's advisor for artificial intelligence and cryptocurrency, said on All-In (2020–present) that Anthropic was among several "AI doomers" that support regulation he saw as overly restrictive. According to The Wall Street Journal, officials close to Sacks examined whether Anthropic's Claude was a "woke AI"; in July, Trump signed an executive order "Preventing Woke AI in the Federal Government ".[15]

Sacks viewed Amodei's decision to attend the World Economic Forum over Trump's second inauguration; his hiring of Biden officials; and Anthropic's association with the philanthropic initiative Open Philanthropy as evidence that Anthropic would not support Trump's agenda.[15] In October 2025, Sacks stated that Anthropic was "running a sophisticated regulatory capture strategy based on fear-mongering."[16] That month, Amodei published a blog post rebuffing "inaccurate claims" from the Trump administration on Anthropic's policies, intensifying the dispute. Amodei's statement included views explicitly espoused by vice president JD Vance.[17] In December, Amodei met with Trump officials and several senators in an effort to improve Anthropic's relationship with the Trump administration.[18]

Dispute

In December 2025, secretary of defense Pete Hegseth announced GenAI.mil, an artificial intelligence platform for the Department of Defense. The department initially contracted Google Gemini for the platform, then OpenAI's ChatGPT.[19][20] The following month, Hegseth announced that the Department of Defense would additionally contract xAI's Grok for use in the military, decrying "woke AI".[21]

In January 2026, Semafor reported that the Department of Defense had conflicted with Anthropic over its policies on lethal military force and that Hegseth's comment on woke AI was a reference to Anthropic.[22] According to Reuters, Anthropic representatives opposed the use of the company's products for surveillance or to develop lethal autonomous weapons.[23] The dispute between Anthropic and the Department of Defense resulted in the termination of a contract worth an estimated US$200 million.[24]

In February 2026, Emil Michael, the under secretary of defense for research and engineering, stated that the Department of Defense would expand access to commercial artificial intelligence systems, including Anthropic's Claude, to unclassified and classified domains.[25] That month, Axios reported that the Department of Defense had used Claude in the United States intervention in Venezuela. Anthropic told Axios that it would reassess its partnership with the Department of Defense after the revelations.[26] After Anthropic refused to agree to allow the Department of Defense to use Claude for "all lawful purposes", the department threatened to cancel its contracts with the company.[27] Hegseth additionally moved to label Anthropic a "supply chain risk", which would have forced military contractors to cut ties with Anthropic.[28] A federal judge blocked most of this designation, describing it as punitive.[29][30] The D.C. Circuit denied Anthropic's emergency motion for a stay of the FASCSA designation in April so it remains in effect for covered systems.[31]

Michael told reporters that Anthropic should "cross the Rubicon" and allow the Department of Defense to dictate the terms of how its technology is used.[32] The position of the Department of Defense, and its tactics during the dispute, were widely criticized on grounds including violating the principles of rule-of-law, market independence and national security.[33][34]

In a June 2026 Bloomberg interview about Claude’s reported role in U.S. military targeting systems, Amodei said Anthropic did not know whether Claude had been used in connection with the Minab school strike, but that, if it had, the use case would not violate Anthropic’s red lines.[35][36] The exchange concerned the 2026 Minab school strike, which Amnesty International described as an unlawful U.S. strike that killed 156 people, including 120 children, and which Human Rights Watch said should be investigated as a war crime.[37][38]

Impact

The dispute caused 1789 Capital, a venture capital firm associated with Donald Trump Jr., to abandon an investment in Anthropic worth hundreds of millions of dollars.[39]

Following the government's actions against Anthropic, OpenAI "rushed",[40] hours before the US started the 2026 Iran war,[41] to get a deal without the constraints that Anthropic had sought.[42]

As of late April, notwithstanding the ND Cal. injunction, "DoW contract cancellations proceed, removal of Claude from DoW systems continues on a 180-day timeline, and Anthropic cannot be used as a prime contractor or subcontractor on DoW covered systems."[31]

Lawsuits

<tbody> </tbody>
Anthropic PBC v. Department of War
CourtNorthern District of California
StartedMarch 9, 2026
Docket nos.3:26-cv-01996
Court membership
Judge sittingRita F. Lin

In March 2026, a federal judge in California, Rita F. Lin, granted a preliminary injunction against the government.[43] Lin wrote:[44]

The Department of War’s records show that it designated Anthropic as a supply chain risk because of its "hostile manner through the press." Punishing Anthropic for bringing public scrutiny to the government’s contracting position is classic illegal First Amendment retaliation. (...) At bottom, Anthropic has shown that these broad punitive measures were likely unlawful and that it is suffering irreparable harm from them. Numerous amici have also described wide-ranging harm to the public interest, including the chilling of open discussion about important topics in AI safety.

Lin ruled in August 2026 that the blacklisting was illegal.[45]

<tbody> </tbody>
Anthropic PBC v. United States Department of War
CourtUnited States Court of Appeals for the District of Columbia Circuit
StartedMarch 9, 2026; 6 months ago (2026-03-09)
Docket nos.26-1049

In April 2026, the Court of Appeals for the D.C. Circuit in a per curiam order denied Anthropic's motion to lift the FASCSA designation.[46] The April order is not final. The court's order said lifting the designation "would force the United States military to prolong its dealings with an unwanted vendor of critical AI services in the middle of a significant ongoing military conflict".[47] According to Wired, "Several experts in government contracting and corporate rights" said "Anthropic has a strong case against the government, but the courts sometimes refuse to overrule the White House on matters related to national security."[48]

See also

References

  1. ^ Markoff, John (November 11, 2014). "Fearing Bombs That Can Pick Whom to Kill". The New York Times. Retrieved February 27, 2026.
  2. ^ Markoff, John (February 28, 2016). "Report Cites Dangers of Autonomous Weapons". The New York Times. Retrieved February 27, 2026.
  3. ^ Markoff, John (May 11, 2016). "Pentagon Turns to Silicon Valley for Edge in Artificial Intelligence". The New York Times. Retrieved February 27, 2026.
  4. ^ Shane, Scott; Metz, Cade; Wakabayashi, Daisuke (May 30, 2018). "How a Pentagon Contract Became an Identity Crisis for Google". The New York Times. Retrieved February 27, 2026.
  5. ^ Davalos, Jackie; Ghaffary, Shirin (January 22, 2025). "Anthropic CEO Says OpenAI's 'Stargate' Venture Seems 'Chaotic'". Bloomberg News. Retrieved March 9, 2026.
  6. ^ Albergotti, Reed (May 30, 2025). "Anthropic emerges as an adversary to Trump's big bill". Semafor. Retrieved March 9, 2026.
  7. ^ Fortinsky, Sarah (June 5, 2025). "Anthropic CEO: GOP AI regulation proposal 'too blunt'". The Hill. Retrieved March 9, 2026.
  8. ^ Heilweil, Rebecca (November 20, 2024). "Anthropic eyes FedRAMP accreditation in quest to sell more AI to government". FedScoop. Retrieved March 9, 2026.
  9. ^ Heilweil, Rebecca (May 22, 2025). "Homeland Security cuts off access to ChatGPT and other commercial AI". FedScoop. Retrieved March 9, 2026.
  10. ^ Heilweil, Rebecca (February 27, 2026). "Trump is booting Anthropic from the military. Palantir helped bring it there". Fast Company. Retrieved March 9, 2026.
  11. ^ "Claude Gov models for U.S. national security customers". Anthropic. June 6, 2025. Retrieved April 29, 2026.
  12. ^ Heilweil, Rebecca (June 5, 2025). "Anthropic drops Claude Gov for natsec customers, hastening the public sector AI race". FedScoop. Retrieved March 9, 2026.
  13. ^ Albergotti, Reed (September 17, 2025). "Anthropic irks White House with limits on models' use". Semafor. Retrieved March 9, 2026.
  14. ^ Shapero, Julia (September 15, 2025). "Anthropic CEO criticizes administration's chip policy, calls for stronger restrictions". The Hill. Retrieved March 9, 2026.
  15. ^ a b Jin, Berber; Ramkumar, Amrith (September 18, 2025). "A Tech CEO's Lonely Fight Against Trump". The Wall Street Journal. Retrieved March 9, 2026.
  16. ^ Curi, Maria (October 21, 2025). "Anthropic tries to defuse White House backlash". Axios. Retrieved March 9, 2026.
  17. ^ Heilweil, Rebecca (October 22, 2025). "Anthropic's relationship with the U.S. government is getting complicated". Fast Company. Retrieved March 9, 2026.
  18. ^ Jones, Rachyl (December 5, 2025). "Anthropic's Dario Amodei on relationship-repair tour in DC". Semafor. Retrieved March 9, 2026.
  19. ^ Sabin, Sam; Demarest, Colin (December 9, 2025). "U.S. military to use Google Gemini for new AI platform". Axios. Retrieved February 27, 2026.
  20. ^ "GenAI.mil's Rapid Expansion Continues With OpenAI Partnership".
  21. ^ "US-Verteidigungsminister will KI-Chatbot Grok für das Militär nutzen" [U.S. secretary of defense wants to use AI chatbot Grok for the military]. Die Zeit (in German). January 13, 2026. Retrieved March 9, 2026.
  22. ^ Albergotti, Reed (January 16, 2026). "Defense Secretary Pete Hegseth jabs Anthropic over safety policies". Semafor. Retrieved March 9, 2026.
  23. ^ Seetharaman, Deepa; Jeans, David; Dastin, Jeffrey (January 29, 2026). "Pentagon clashes with Anthropic over military AI use, sources say". Reuters. Retrieved March 9, 2026.
  24. ^ "What to know about the clash between the Pentagon and Anthropic over military's AI use". AP News. February 28, 2026. Retrieved May 22, 2026.
  25. ^ Jeans, David; Seetharaman, Deepa (February 11, 2026). "Pentagon pushing AI companies to expand on classified networks, sources say". Reuters. Retrieved March 9, 2026.
  26. ^ Lawler, Dave; Curi, Maria (February 13, 2026). "Pentagon's use of Claude during Maduro raid sparks Anthropic feud". Axios. Retrieved March 9, 2026.
  27. ^ Lawler, Dave; Curi, Maria (February 14, 2026). "Pentagon threatens to cut off Anthropic in AI safeguards dispute". Axios. Retrieved March 9, 2026.
  28. ^ Lawler, Dave; Curi, Maria; Allen, Mike (February 16, 2026). "Pentagon threatens Anthropic punishment". Axios. Retrieved March 9, 2026.
  29. ^ O'Donnell, James, Pentagon’s culture war tactic against Anthropic has backfired: Decisions to tweet first and lawyer later didn’t sit well with a federal judge, who last week halted the government’s punishment of the AI company., MIT Technology Review, March 30, 2026
  30. ^ Cole, Hadas Gold, Devan (March 26, 2026). "Judge blocks Pentagon's effort to 'punish' Anthropic by labeling it a supply chain risk | CNN Business". CNN. Retrieved May 22, 2026.{{cite web}}: CS1 maint: multiple names: authors list (link)
  31. ^ a b Lee, Andrew R.; Ramsden, Michelle; Loring, Jason M.; Ryan, Graham H. (April 27, 2026). "Two Courts, Two Postures: What the DC Circuit's Stay Denial Means for the Anthropic-Pentagon Dispute". Jones Walker. Retrieved June 9, 2026. … the actual state of play is narrower than the headlines suggest. Inside the DoW, the § 4713 designation remains operative. … DoW contract cancellations proceed, removal of Claude from DoW systems continues on a 180-day timeline, and Anthropic cannot be used as a prime contractor or subcontractor on DoW covered systems.
  32. ^ Vincent, Brandi (February 19, 2026). "Pentagon CTO urges Anthropic to 'cross the Rubicon' on military AI use cases amid ethics dispute". DefenseScoop. Retrieved March 10, 2026.
  33. ^ Sage, Emma Isabella (February 25, 2026). "The Pentagon's Dangerous Adoption of AI". The Dispatch. Retrieved May 22, 2026.
  34. ^ Sylvain, Olivier. "I've been studying Big Tech for a long time. What just happened with Anthropic and the Pentagon terrifies me". Fortune. Retrieved May 22, 2026.
  35. ^ "Watch Inside Anthropic, the $965 Billion AI Titan". Bloomberg. June 10, 2026. Retrieved June 18, 2026.
  36. ^ "Bloomberg Business on X: "On The Circuit with Emily Chang, Anthropic CEO Dario Amodei addresses reports that the US military is using Claude as part of the US-Israel war in Iran"". X. Bloomberg. June 10, 2026. Retrieved June 18, 2026.
  37. ^ "USA/Iran: Those responsible for deadly and unlawful US strike on school that killed over 100 children must be held accountable". Amnesty International. March 16, 2026. Retrieved June 18, 2026.
  38. ^ "US/Israel: Investigate Iran School Attack as a War Crime". Human Rights Watch. March 7, 2026. Retrieved June 18, 2026.
  39. ^ Hagey, Keach; Ramkumar, Amrith; Acosta, Deborah; Bergengruen, Vera (February 17, 2026). "'Woke' AI Feud Escalates Between Pentagon and Anthropic". The Wall Street Journal. Retrieved March 10, 2026.
  40. ^ @sama (March 2, 2026). "Tweet". Twitter. 5. One thing I think I did wrong: we shouldn't have rushed to get this out on Friday. The issues are super complex, and demand clear communication. We were genuinely trying to de-escalate things and avoid a much worse outcome, but I think it just looked opportunistic and sloppy. Good learning experience for me as we face higher-stakes decisions in the future.
  41. ^ Capoot, Ashley (March 3, 2026). "Sam Altman tells OpenAI staffers that military's 'operational decisions' are up to the government". CNBC. The meeting occurred four days after OpenAI announced its DOD arrangement, which landed just hours before the U.S. and Israel began carrying out strikes against Iran.
  42. ^ O'Donnell, James (March 2, 2026). "OpenAI's "compromise" with the Pentagon is what Anthropic feared". MIT Technology Review. However, the published excerpt "does not give OpenAI an Anthropic-style, free-standing right to prohibit otherwise-lawful government use," wrote Jessica Tillipman, associate dean for government procurement law studies at George Washington University's law school. It simply states that the Pentagon can't use OpenAI's tech to break any of those laws and policies as they're stated today. The whole reason Anthropic earned so many supporters in its fight—including some of OpenAI's own employees—is that they don't believe these rules are good enough to prevent the creation of AI-enabled autonomous weapons or mass surveillance. And an assumption that federal agencies won't break the law is little assurance to anyone who remembers that the surveillance practices exposed by Edward Snowden had been deemed legal by internal agencies and were ruled unlawful only after drawn-out battles (not to mention the many surveillance tactics allowed under current law that AI could expand). On this front, we've essentially ended up back where we started: allowing the Pentagon to use its AI for any lawful use.
  43. ^ Dave, Paresh (March 26, 2026). "Anthropic Supply-Chain-Risk Designation Halted by Judge". Wired.com.
  44. ^ "ORDER GRANTING MOTION FOR PRELIMINARY INJUNCTION" (PDF). CourtListener. March 26, 2026.
  45. ^ Capoot, Ashley (August 28, 2026). "Judge blocks Pentagon blacklist of Anthropic as supply chain risk". CNBC. Retrieved September 19, 2026.
  46. ^ "Judge rejects DOJ bid to delay Anthropic appeal in Pentagon dispute". POLITICO. April 24, 2026. Retrieved April 29, 2026.
  47. ^ Anthropic PBC v. United States Department of War, 2026 WL 1042493 (DC Cir. April 8, 2026).
  48. ^ Dave, Paresh (April 8, 2026). "Anthropic vs. the Pentagon: This Time, a US Court Has Ruled in the Government's Favor". Wired.com. Archived from the original on April 8, 2026.

Further reading

Retrieved from "https://deeppolitics.org/index.php?title=Anthropic–United_States_Department_of_Defense_dispute&oldid=71593"