The Information Machine

Pentagon and Microsoft Pulling Back from Anthropic Claude · history

Version 6

2026-05-25 19:10 UTC · 174 items

What

The U.S. Department of War has designated Anthropic a 'Supply Chain Risk,' triggering a legal and commercial standoff that splits both the federal judiciary and the executive branch itself [11][9]. D.C. courts upheld the designation while California courts twice blocked broader contracting restrictions—both California orders now under Pentagon appeal [15]. Within the same administration, the White House cleared a $125 million Anthropic-NSA deal over direct Pentagon objection [22]. OpenAI announced a Pentagon deal within days of the Anthropic ban, stepping into the competitive gap [6]. The Congressional Research Service has now formally examined the dispute [25], signaling congressional attention to a conflict existing procurement law was not designed to resolve.

Why it matters

The intra-executive contradiction—one arm of the U.S. government excluding Anthropic on national security grounds while another signs an intelligence contract with the same company—tests whether contradictory executive procurement policies toward a single vendor can coexist without a legal resolution mechanism. OpenAI's entry as a direct beneficiary [6] and the CRS's formal examination [25] add a competitive-norms question and a legislative-attention dimension: whether the AI company that accepts the Pentagon's autonomous weapons terms will out-compete the one that refuses them, and whether Congress will eventually intervene.

Open questions

  • If the Pentagon successfully appeals the California orders [15], Anthropic would lose all active judicial protection — does the $125 million NSA deal [22] provide meaningful durable access to federal revenue as a substitute?

  • The CRS has formally examined the dispute [25] — has Congress begun moving toward any legislative response, such as procurement reform or statutory limits on Supply Chain Risk designations applied to AI vendors?

  • OpenAI announced a Pentagon deal within days of the Anthropic ban [6] — do its terms include the autonomous weapons and surveillance use cases Anthropic refused, and does this create competitive pressure on Anthropic's ethical stance?

  • The EFF argues that privacy protections dependent on individual corporate decisions are structurally insufficient [32] — what systemic legislative or regulatory mechanisms are being proposed, and have any legislators engaged publicly?

Narrative

The conflict traces to a blunt ultimatum over AI in military contexts. In late February 2026, Anthropic CEO Dario Amodei publicly refused Pentagon demands that Claude be made available for autonomous weapons development and mass surveillance [1][2][3]. Trump ordered the federal government to stop using Anthropic's tools [4], and the Department of Defense—operating as the 'Department of War' under the Trump administration, confirmed by the agency's own war.gov domain [5]—responded by designating Anthropic a 'Supply Chain Risk,' a national security procurement classification that bars a company from classified government contracts. Within days, OpenAI announced a Pentagon deal, stepping directly into the commercial opening the ban created [6]. Anthropic published its own account of the dispute titled 'Where things stand with the Department of War' [7], framing its refusal as a principled position rather than a negotiating posture.

The legal battle that followed split the federal judiciary by jurisdiction. California courts granted Anthropic a preliminary injunction in late March 2026 [8] and a broader blocking order in May [9], while D.C. courts upheld the Supply Chain Risk designation at both the district and circuit levels [10][11]. A bipartisan coalition of 149 former federal and state judges filed an amicus brief supporting Anthropic [12], joined by Big Law partners [13] and the Society for the Rule of Law [14]—an unusual concentration of cross-partisan legal-establishment backing for a commercial AI vendor. The Pentagon is now appealing the California orders [15], which would eliminate Anthropic's only active judicial protection. Multiple law firms—Mayer Brown, Fluet Law, Pearl Cohen, and Goodwin Law—have published detailed contractor guidance on the practical effects of the designation, treating the case as a landmark in AI procurement law [16][17][18][19].

The most structurally novel dimension of the dispute is a documented intra-executive contradiction. The Department of War expanded its classified AI vendor roster to seven or eight major tech companies that explicitly exclude Anthropic [20][5] and is evaluating rival models as direct Claude replacements [21], while the White House cleared a classified $125 million Anthropic-NSA arrangement over direct Pentagon objection [22], framed within a $9 billion commitment to advance spy-agency AI capabilities [23]. Trump said a DoD deal is 'possible' [24], leaving the contradiction formally unresolved. The Congressional Research Service published a formal report on the dispute [25], signaling that Congress is now officially tracking a conflict that existing procurement law was not designed to arbitrate.

The commercial consequences extend beyond direct government contracts. Private-sector defense tech companies began dropping Claude after the blacklist [26], and Microsoft canceled internal Claude Code developer licenses, redirecting engineers to GitHub Copilot [27][28]. Despite these losses, Anthropic is on track for $10.9 billion in Q2 2026 revenue and projected to reach its first profitable quarter [29]. Analysts disagree on what the standoff means: The Atlantic frames Anthropic's refusal as a stand that may prove strategically and ethically vindicating [30], while Brookings asks whether responsible AI commitments can survive military procurement demands industry-wide [31]. The EFF raises a structural critique that transcends the specific actors: privacy protections that depend on the decisions of individual corporate leaders are inherently insufficient regardless of their intentions, and systemic safeguards are needed [32].

Timeline

  • 2026-02: Defense One and Scientific American report replacing Anthropic's Pentagon tools would take months and would not be straightforward. [41][42]
  • 2026-02-26: Anthropic CEO Dario Amodei publicly refuses Pentagon demands to permit Claude's use in autonomous weapons development and mass surveillance. [1][2][3]
  • 2026-02: Trump orders the U.S. government to stop using Anthropic's tools after Anthropic refuses the weapons and surveillance ultimatum. [4]
  • 2026-02-27: OpenAI announces a Pentagon deal days after the Anthropic ban, stepping into the direct competitive gap. [6]
  • 2026-03-04: CNBC reports private-sector defense tech companies are dropping Claude following the Pentagon blacklist. [26]
  • 2026-03: Anthropic announces a $200M DoD responsible AI agreement, challenges the Supply Chain Risk designation in court, and publishes 'Where things stand with the Department of War.' [34][43][7]
  • 2026-03-18: A bipartisan coalition of 149 former federal and state judges files an amicus brief supporting Anthropic's suit against the Department of Defense. [12]
  • 2026-03-25: Big Law partners and the Society for the Rule of Law file amicus briefs in the D.C. Circuit in support of Anthropic. [13][14]
  • 2026-03-26: Anthropic wins a preliminary injunction in California federal court blocking initial restrictions on its federal contracting access. [8]
  • 2026-04-08: A D.C. district court denies Anthropic's motion to lift the Department of War's Supply Chain Risk designation. [10][36]
  • 2026-04-21: Trump says an Anthropic deal for Department of Defense use is 'possible.' [24]
  • 2026-05-01: The Pentagon strikes classified AI deals with seven to eight major tech companies, explicitly excluding Anthropic, confirmed in the DoW's official press release. [20][44][5]
  • 2026-05-19: D.C. Circuit judges publicly question the legality of the DoD's move to bar Anthropic from government contracts during appellate proceedings. [45][46]
  • 2026-05-20: CNBC reports Anthropic is on track for $10.9 billion in Q2 2026 revenue, projecting its first profitable quarter. [29]
  • 2026-05-21: A California federal court broadly blocks the Trump administration from restricting Anthropic's federal contracts; the Pentagon is reported testing rival models; Microsoft cancels internal Claude Code developer licenses. [9][21][27][28]
  • 2026-05-22: The D.C. Circuit upholds the Pentagon's Supply Chain Risk designation against Anthropic; the White House approves $9 billion for spy-agency AI. [33][11][23]
  • 2026-05: The Pentagon appeals the California court orders that had blocked restrictions on Anthropic's federal contracting. [15]
  • 2026-05-24: The White House and Anthropic finalize a $125 million deal for U.S. spy agencies to access Claude, cleared over direct Pentagon objection. [35][47][22]
  • 2026-05: The Congressional Research Service publishes a formal report on the Pentagon-Anthropic dispute over autonomous weapon systems. [25]
  • 2026-05: Pearl Cohen and Goodwin Law publish federal contractor guidance on the Supply Chain Risk designation's practical effects, joining Mayer Brown and Fluet Law in treating the case as a procurement law landmark. [18][19]

Perspectives

U.S. Department of War

Demanded Anthropic modify usage terms to permit weapons and surveillance applications; after refusal, designated Anthropic a Supply Chain Risk and prevailed at both D.C. district and circuit levels; expanded classified AI work to companies that explicitly exclude Anthropic; now appealing the California orders that had partially protected Anthropic's contracting access.

Evolution: Consistent with prior synthesis; the Pentagon's appeal of the California rulings [17852] continues to be the key active procedural move.

[4][10][33][20][5][15][11]

Anthropic

Refused the Pentagon's weapons and surveillance ultimatum as a matter of policy; lost at both D.C. court levels; holds a California preliminary injunction now under Pentagon appeal; published a primary-source public statement on the dispute; holds a $200M DoD agreement and is finalizing a $125M White House intelligence community deal.

Evolution: Consistent with prior synthesis; Pearl Cohen and Goodwin Law now join the law firms treating Anthropic's case as a landmark [20286][20316], broadening the legal-guidance ecosystem documenting its position.

[2][1][3][7][34][35][22][8]

Federal judiciary

Split across jurisdictions — California granted two blocking orders [19760][10911] while D.C. courts upheld the designation at both levels [13049][19759]; the Pentagon is appealing the California orders [17852], which could eliminate Anthropic's only active judicial protection.

Evolution: Consistent with prior synthesis.

[9][8][10][33][36][15][11]

Trump administration / White House

Trump ordered the government-wide Anthropic ban after the weapons ultimatum refusal but has also said a DoD deal is 'possible'; the White House separately cleared a $125M Anthropic-NSA deal over direct Pentagon objection and committed $9 billion to spy-agency AI, creating a documented intra-executive contradiction.

Evolution: Consistent with prior synthesis.

[4][24][35][23][22]

Legal community (former judges, Big Law, law firm analysts)

149 former judges, Big Law partners, and the Society for the Rule of Law filed amicus briefs supporting Anthropic; separately, Mayer Brown, Fluet Law, Pearl Cohen, and Goodwin Law have published contractor guidance framing the case as a landmark in AI procurement law.

Evolution: Pearl Cohen [20286] and Goodwin Law [20316] join the law firm analyst voice previously limited to Mayer Brown and Fluet Law, deepening the documented legal-guidance ecosystem.

[12][13][14][16][17][18][19]

EFF

The Pentagon-Anthropic conflict reveals a structural problem: privacy protections that depend on the decisions of individual corporate leaders are inherently insufficient regardless of intent, and systemic safeguards are needed rather than reliance on corporate virtue.

Evolution: Item 20317 reconfirms the EFF's position without new claims; stance unchanged from prior synthesis.

[37][32]

OpenAI

Announced a Pentagon deal within days of the Anthropic ban, positioning itself as the direct commercial beneficiary of the competitive opening the Supply Chain Risk designation created.

Evolution: First appearance as a named perspective; NPR [5890] documents the timing and competitive positioning.

[6]

Microsoft

Canceled internal Claude Code developer licenses and directed employees to GitHub Copilot, a move attributed to financial consolidation around Microsoft's own tooling; Claude remains accessible through Copilot CLI and powers features in Microsoft 365.

Evolution: Consistent with prior synthesis; additional confirming sources reinforce the cancellation story without changing the reported rationale.

[27][38][39][28]

Tensions

  • Department of War vs. Anthropic on autonomous weapons use: the DoW demands Claude be deployable for weapons development and mass surveillance; Anthropic's policies prohibit this and CEO Amodei has publicly refused to change them [5869], triggering the Supply Chain Risk designation and all subsequent legal and procurement consequences. [2][1][3][4][7]
  • D.C. courts vs. California courts on Anthropic's legal status: California granted two blocking orders [19760][10911]; D.C. upheld the designation at both levels [13049][19759]; the Pentagon appeals the California orders [17852], attempting to close the one judicial gap in Anthropic's favor. [9][8][10][33][15][11]
  • Pentagon exclusion vs. White House intelligence inclusion: the DoW bars Anthropic and signs AI deals with companies that explicitly exclude it [18433], while the White House clears a $125M NSA deal over direct Pentagon objection [18431] — two parts of the same executive branch pursuing contradictory procurement policies toward the same vendor. [20][5][35][22][23]
  • EFF structural critique vs. Anthropic ethical-stand framing: The Atlantic and Anthropic frame the CEO's refusal as a principled stand that may prove vindicating [10910][5869]; the EFF argues that relying on individual corporate decisions to protect privacy is structurally insufficient regardless of actor intentions, and that systemic safeguards are needed [20317]. [30][7][32]
  • The Atlantic vs. Brookings on industry-wide implications: The Atlantic frames Anthropic's refusal as potentially strategic and vindicating; Brookings asks whether the feud signals that responsible AI commitments cannot survive contact with military procurement — opposite readings of what the standoff means for the broader AI industry. [30][40][31]
  • OpenAI vs. Anthropic on terms of Pentagon engagement: OpenAI signed a Pentagon deal within days of the Anthropic ban [5890], raising the unresolved question of whether it accepted the autonomous weapons and surveillance terms Anthropic refused — a competitive dynamic that could pressure Anthropic's ethical stance over time. [6][2][1]

Sources

  1. [1] Anthropic rejects Pentagon terms for lethal use of its chatbot Claude — reactive:anthropic-enterprise-losses
  2. [2] Anthropic refuses Pentagon's new terms, standing firm on lethal ... — reactive:anthropic-enterprise-losses
  3. [3] Anthropic boss rejects Pentagon demand to drop AI safeguards — reactive:anthropic-ai-values-widening
  4. [4] Anthropic wanted the Pentagon to agree not to use its AI for autonomous weapons and mass surveillance. So Trump ordered the government to stop using it altogether. — reactive:anthropic-enterprise-losses
  5. [5] Classified Networks AI Agreements - U.S. Department of War — reactive:anthropic-enterprise-losses
  6. [6] OpenAI announces Pentagon deal after Trump bans Anthropic - NPR — reactive:openai-advanced-account-security
  7. [7] Where things stand with the Department of War - Anthropic — reactive:openai-financial-strategy
  8. [8] Anthropic wins preliminary injunction in Trump DOD fight — reactive:anthropic-enterprise-losses
  9. [9] Judge blocks Trump administration from limiting Anthropic's contracts with federal government — reactive:anthropic-enterprise-losses
  10. [10] Federal Court Denies Anthropic's Motion to Lift 'Supply ... — reactive:anthropic-enterprise-losses
  11. [11] Appeals court decides against Anthropic in latest round of its AI battle with the Trump administration | PBS News — reactive:anthropic-enterprise-losses
  12. [12] BIPARTISAN COALITION OF 149 FORMER FEDERAL AND STATE JUDGES FILES BRIEF SUPPORTING ANTHROPIC’S SUIT AGAINST THE DEPARTMENT OF DEFENSE — reactive:anthropic-enterprise-losses
  13. [13] Big Law Partners File Amicus Briefs in Support of Anthropic in Row ... — reactive:anthropic-enterprise-losses
  14. [14] Amicus Brief in Anthropic PBC vs. Department of War, D.C. Circuit Court of Appeals - Society for the Rule of Law — reactive:anthropic-enterprise-losses
  15. [15] Pentagon appealing order to remove Anthropic 'supply chain risk' label — reactive:anthropic-enterprise-losses
  16. [16] Anthropic Supply Chain Risk Designation Takes Effect - Mayer Brown — reactive:anthropic-enterprise-losses
  17. [17] The Department's War with Anthropic: Litigation Update - Fluet Law — reactive:anthropic-enterprise-losses
  18. [18] Anthropic Sues Department of Defense Over Supply Chain Risk Designation - Pearl Cohen — reactive:openai-corporate-transition
  19. [19] Is Claude a Supply Chain Risk? What Federal Contractors Need to Know About This Designation | Insights & Resources | Goodwin — reactive:anthropic-enterprise-losses
  20. [20] DOD expands its classified AI work with 8 companies — excluding Anthropic — amid ongoing dispute | DefenseScoop — reactive:anthropic-enterprise-losses
  21. [21] Pentagon Tests Rival AI Models in Race to Replace ... — reactive:anthropic-enterprise-losses
  22. [22] White House Clears Anthropic NSA Deal Over Pentagon Objection | AI Weekly — reactive:anthropic-enterprise-losses
  23. [23] White House Approves $9 Billion for Spy Agencies to Catch Up on AI - GV Wire — reactive:anthropic-enterprise-losses
  24. [24] Trump says Anthropic deal is 'possible' for Department of Defense use — reactive:anthropic-enterprise-losses
  25. [25] Pentagon-Anthropic Dispute over Autonomous Weapon Systems — reactive:anthropic-enterprise-losses
  26. [26] Defense tech companies are dropping Claude after Pentagon's Anthropic blacklist — reactive:anthropic-enterprise-losses
  27. [27] Microsoft is ditching Claude Code for Copilot CLI - Windows Central — reactive:anthropic-enterprise-losses
  28. [28] Microsoft starts canceling Claude Code licenses — reactive:anthropic-enterprise-losses
  29. [29] Anthropic set to hit $10.9 billion in revenue in Q2, source says - CNBC — reactive:anthropic-enterprise-losses
  30. [30] Anthropic’s Ethical Stand Could Be Paying Off - The Atlantic — reactive:anthropic-enterprise-losses
  31. [31] Does the Anthropic–Pentagon feud mean the end of responsible AI? — reactive:anthropic-enterprise-losses
  32. [32] Your Privacy Shouldn't Be A Corporate Decision — reactive:anthropic-enterprise-losses
  33. [33] Federal Appeals Court Upholds Pentagon's Supply-Chain ... — reactive:anthropic-enterprise-losses
  34. [34] Anthropic awarded $200M DOD agreement for AI capabilities \ Anthropic — reactive:anthropic-enterprise-losses
  35. [35] 🤯 ANTHROPIC just secured a $125M White House Intel deal. — reactive:anthropic-enterprise-losses (2026-05-24)
  36. [36] Anthropic loses appeals court bid to temporarily block DOD ruling — reactive:anthropic-ai-values-widening
  37. [37] The Anthropic-DOD Conflict: Privacy Protections Shouldn’t Depend On the Decisions of a Few Powerful People | Electronic Frontier Foundation — reactive:openai-financial-strategy
  38. [38] Microsoft pulls Claude Code licenses and pushes developers back ... — reactive:anthropic-enterprise-losses
  39. [39] Microsoft starts canceling Claude Code licenses — reactive:anthropic-enterprise-losses
  40. [40] Donald Trump Declares War on Anthropic - The Atlantic — reactive:anthropic-enterprise-losses
  41. [41] It would take the Pentagon months to replace Anthropic’s AI tools: sources - Defense One — reactive:anthropic-enterprise-losses
  42. [42] Why replacing Anthropic at the Pentagon could take months | Scientific American — reactive:anthropic-enterprise-losses
  43. [43] Anthropic Challenges DoW’s Supply Chain Risk Designation — reactive:anthropic-enterprise-losses
  44. [44] Pentagon inks deals with seven AI companies for classified military work — reactive:openai-microsoft-partnership-amendment
  45. [45] Potential splits emerged between D.C. Circuit judges questioning the legality of the DOD's move to bar Anthropic from go... — reactive:anthropic-enterprise-losses (2026-05-20)
  46. [46] Potential splits emerged between D.C. Circuit judges questioning the legality of the DOD's move to bar Anthropic from go... — reactive:anthropic-enterprise-losses (2026-05-19)
  47. [47] White House Nears Deal with Anthropic for AI Use in Intelligence — reactive:anthropic-enterprise-losses