Status update as of June 2026
COURT BLOCKED IT: The Supply Chain Risk Designation Is Preliminarily Enjoined, Not in Effect
The "supply chain risk" designation described further down this page is not currently in force. On March 26, 2026, a federal judge granted Anthropic a preliminary injunction and blocked the designation while the case proceeds. Read the rest of this page as the background to an ongoing dispute, not as a description of a ban that is operating today.
1
Preliminary injunction granted March 26, 2026: Judge Rita F. Lin of the U.S. District Court for the Northern District of California granted Anthropic a preliminary injunction, blocking the 10 U.S.C. § 3252 designation, the presidential directive, and the directive issued by Defense Secretary Pete Hegseth. The court found Anthropic likely to succeed on its First Amendment retaliation, Fifth Amendment due process, and Administrative Procedure Act (arbitrary and capricious) claims. (FindLaw; Mayer Brown.)
2
Litigation is ongoing: A preliminary injunction is not a final judgment. It keeps the designation and the related directives blocked while the case proceeds to the merits. The government may appeal or continue to litigate, so the legal status can still change.
3
Parties reported back at the table: By later in the spring, the Pentagon had announced AI deals with several large technology companies, and Anthropic and the Pentagon were reported to be back in negotiations. (CNN, May 1, 2026.)
How to read what follows. The banners and cards below this point describe the threat, the designation, and the early lawsuit as they unfolded in late February and March 2026. I have kept that history because the underlying events are real and well sourced, but the designation itself has since been blocked by the court ruling summarized above. Where a card describes the ban as in effect, treat it as a snapshot of that moment, superseded by this status update.
March 9 to March 26, 2026, the lawsuit and the injunction
LAWSUIT, THEN INJUNCTION: Anthropic Sued the Pentagon, and a Court Blocked the Designation
On March 9, 2026, Anthropic sued the Department of Defense over the supply chain risk designation, according to Axios. On March 26, 2026, the court granted a preliminary injunction blocking the designation, as reported by FindLaw and Mayer Brown.
1
The lawsuit: On March 9, 2026, Anthropic sued the Pentagon, challenging the supply chain risk designation. (Axios.)
2
The constitutional and statutory claims: Anthropic's suit raised First Amendment retaliation, Fifth Amendment due process, and Administrative Procedure Act (arbitrary and capricious) theories, arguing the designation was retaliation for its policy positions and was imposed without adequate process. (FindLaw; Mayer Brown.)
3
The injunction: On March 26, 2026, Judge Rita F. Lin (N.D. Cal.) granted a preliminary injunction blocking the § 3252 designation, the presidential directive, and the Hegseth directive, finding Anthropic likely to succeed on its First Amendment, Fifth Amendment, and APA claims. (FindLaw; Mayer Brown.)
4
What the company said about the stakes: Anthropic argued the designation threatened substantial revenue because it reached beyond direct government contracts and pressured defense contractors over their commercial use of Claude. As I read the public reporting, that broad reach is exactly what made the harm plausible enough to support emergency relief. (Mayer Brown.)
5
Framing of the suit: Anthropic framed the case as a challenge to the government blacklisting a company over policy disagreements, rather than an attempt to force the government to buy its product. The core legal question, in my view, is whether a supply chain statute aimed at foreign-adversary risk can be turned against a domestic company for keeping ethical guardrails. (Just Security.)
6
Why it matters beyond Anthropic: The ruling addressed whether the government can condition contracting on a company stripping safety guardrails, and whether doing so by designation, without a hearing, survives First Amendment and due process scrutiny. In my view, the reasoning could influence how every frontier AI vendor negotiates federal terms going forward, though a preliminary injunction sets no final precedent.
February 27, 2026, the designation (since blocked, see status update above)
WHAT THE GOVERNMENT DID ON FEB 27: Anthropic Designated a "Supply Chain Risk" Under 10 U.S.C. 3252
Status note: The actions described in this section were taken on February 27, 2026, but a federal court preliminarily blocked the designation and the related directives on March 26, 2026 (see the green status update at the top). Read what follows as a record of what happened that day, not as a description of measures in force today.
1
Supply chain risk designation: On February 27, 2026, Defense Secretary Pete Hegseth designated Anthropic a "supply chain risk" under 10 U.S.C. § 3252 and the Federal Acquisition Supply Chain Security Act (FASCSA, 2018), a framework typically used to address foreign-adversary supply chain concerns. (Mayer Brown; Just Security.)
2
An unprecedented use of the authority, per commentators: Legal commentators described the move as an unprecedented use of this authority against a U.S. company, a tool more commonly associated with foreign-adversary risk. I treat that as informed commentary rather than a settled historical fact. (Mayer Brown; Just Security.)
3
Direction to federal agencies: On February 27, 2026, the administration directed federal agencies to stop using Anthropic, alongside the supply chain risk designation. The court later blocked the designation, the presidential directive, and the Hegseth directive (see status update above). (Just Security; NBC News.)
4
The two red lines Anthropic would not cross: Anthropic declined updated Pentagon contract terms over two stated red lines: no use of Claude for mass surveillance of Americans, and no fully autonomous weapons without human oversight. (Mayer Brown; NBC News.)
5
OpenAI's Pentagon agreement: Hours after the designation, OpenAI announced a roughly $200 million Pentagon agreement to use its services in classified settings, positioned to take over much of Anthropic's Pentagon business. Sam Altman said he shared Anthropic's red lines, but had been negotiating OpenAI's own deal, which drew criticism. (Yahoo Finance; CNN.)
6
Anthropic's response: Anthropic publicly stated it would not change its position on mass surveillance or fully autonomous weapons and would challenge the designation in court. As reported, the company followed through with its lawsuit on March 9, 2026. (NBC News; Axios.)
7
The contracting reach that worried defense vendors: Government-contracts lawyers flagged that a supply chain risk designation can ripple out to contractors and subcontractors who must account for the designated company in their own supply chains. That ripple effect is the part the court's injunction has, for now, put on hold. (Mayer Brown.)
8
My read: In my view, the core of Anthropic's case was always the mismatch between a foreign-adversary supply chain tool and a domestic policy dispute, and the lack of process before designation. The March 26 preliminary injunction is consistent with that read, though it is preliminary and the merits are not yet decided.
Late February 2026, the background
How This Started: The Pentagon Pressed Anthropic on Contract Terms, and Anthropic Declined
The dispute grew out of Anthropic's role as a Pentagon AI vendor. Claude was the first major AI model deployed on U.S. government classified networks, tied to a roughly $200 million Department of Defense relationship. (Yahoo Finance.)
Anthropic declined updated Pentagon contract terms over two red lines: no use of Claude for mass surveillance of Americans, and no fully autonomous weapons without human oversight. (Mayer Brown; NBC News.) On February 27, 2026, after Anthropic held its position, the government designated the company a supply chain risk. A federal court later blocked that designation on March 26, 2026 (see the status update at the top of this page).
Around the same period, Anthropic published RSP 3.0, replacing its hard safety commitments with nonbinding "Frontier Safety Roadmaps." The company has said the timing relative to the Pentagon dispute was coincidental.
10 USC 3252
Supply Chain Risk
What the Designation Would Have Done (Now Blocked)
Currently enjoined. A federal court preliminarily blocked this designation on March 26, 2026. What follows describes how the designation was supposed to operate when it was issued. (FindLaw; Mayer Brown.)
The designation under 10 U.S.C. § 3252 and FASCSA was meant to flow through the defense procurement system, prompting Department of Defense vendors and contractors to account for the designated company in their own supply chains. Legal commentators described applying this authority to a domestic company as unprecedented. (Mayer Brown; Just Security.)
The legal fight, in my view: Anthropic argued the designation could not dictate how contractors use Claude for non-government customers, and that imposing it as retaliation for safety positions, without a hearing, ran into the First Amendment, the Fifth Amendment, and the APA. The court found those arguments likely to succeed when it granted the preliminary injunction on March 26, 2026. (FindLaw.)
RSP 3.0
Safety Policy
Anthropic Drops Hard Safety Limits
Effective February 24, 2026, Anthropic replaced its Responsible Scaling Policy with RSP Version 3.0. The key change: the categorical commitment to pause training if safety measures weren't proven is gone.
New framework: nonbinding "Frontier Safety Roadmaps" with public progress reports every 3-6 months. The pause trigger now requires both AI race leadership and material catastrophic risk, a much higher bar.
Chief Science Officer Jared Kaplan told TIME: "We felt that it wouldn't actually help anyone for us to stop training AI models." The decision took nearly a year of internal debate.
Industry impact: OpenAI and Google DeepMind adopted similar frameworks after Anthropic's original RSP. A rollback by the originator may reshape what "responsible scaling" means industry-wide.
India Summit
Global Policy
US Opposes Global AI Guardrails
At the AI Impact Summit in New Delhi (February 20, 2026), the Trump administration stood alone opposing centralized regulation of generative AI, blocking a multinational statement on AI governance.
US position: guardrails slow innovation and give adversaries like China an edge. Critics warn a poorly governed race toward AGI poses existential risk. India, EU, and dozens of nations favor caution.
Competitors
Compliance
Where Other AI Companies Stand
OpenAI: Hours after the designation, OpenAI announced a roughly $200 million Pentagon agreement to use its services in classified settings, positioned to replace much of Anthropic's Pentagon business. Sam Altman said he shared Anthropic's red lines, no mass surveillance and no fully autonomous weapons, but had been negotiating OpenAI's own deal, which drew criticism. (Yahoo Finance; CNN.)
Anthropic: Sued the Pentagon on March 9, 2026, and won a preliminary injunction on March 26, 2026 blocking the designation and the related directives. The litigation is ongoing. (Axios; FindLaw.)
The wider field: By later in the spring, the Pentagon had struck AI deals with several large technology companies, and Anthropic and the Pentagon were reported to be back at the negotiating table. (CNN, May 1, 2026.)
Feb 27 Directive
Court Blocked It
The Direction to Federal Agencies, Then the Injunction
On February 27, 2026, the administration directed federal agencies to stop using Anthropic, alongside the supply chain risk designation. (Just Security; NBC News.)
On March 26, 2026, a federal court preliminarily blocked the designation, the presidential directive, and the Hegseth directive, so this direction to agencies is not currently in force. (FindLaw; Mayer Brown.)
Employee Revolt
Industry
The Staff Reaction
Reporting at the time described unease inside other AI labs about the episode, including criticism of OpenAI's decision to pursue its own Pentagon deal while Anthropic was being designated. (CNN.)
In my view, the reaction echoes earlier industry pushback on military AI work, but I would not overstate it: a staff reaction is not the same as a company changing its position.
Where Things Stand
Negotiations
Back at the Negotiating Table
By later in the spring, the Pentagon had struck AI deals with several large technology companies, and Anthropic and the Pentagon were reported to be back in negotiations. (CNN, May 1, 2026.)
In my view, the practical takeaway is that the court order changed the leverage. With the designation blocked, the parties had room to negotiate rather than litigate to a final judgment, though the case is not over.
AI Policy Timeline, 2025 to 2026
Jan 23, 2025
Trump signs EO 14179 revoking Biden AI safety executive order. Directs removal of all "barriers to American AI leadership."
Dec 11, 2025
Trump signs "Ensuring a National Policy Framework for AI" EO. Establishes DOJ AI Litigation Task Force to challenge state AI laws. $42B BEAD broadband funding conditioned on state AI law repeal.
Jan 1, 2026
New state AI laws take effect in 38 states. Colorado AI Act delayed to June 30, 2026 under preemption pressure.
Jan 10, 2026
DOJ AI Litigation Task Force becomes operational. Begins identifying state laws to challenge in federal court.
Feb 20, 2026
AI Impact Summit, New Delhi. US opposes global AI guardrails statement. "Rate Payer Protection Pledge" announced: tech companies must power their own AI data centers.
Late Feb 2026
Anthropic declines updated Pentagon contract terms over two red lines: no mass surveillance of Americans, no fully autonomous weapons without human oversight. (Mayer Brown; NBC News.)
Feb 27, 2026
DESIGNATION: Hegseth designates Anthropic a "supply chain risk" under 10 U.S.C. § 3252 and FASCSA, and the administration directs federal agencies to stop using Anthropic. Commentators call the use against a U.S. company unprecedented. (Mayer Brown; Just Security.)
Feb 27, 2026
OpenAI announces a roughly $200M Pentagon agreement for classified settings, positioned to replace much of Anthropic's Pentagon business. Sam Altman said he shared Anthropic's red lines but had negotiated OpenAI's own deal, which drew criticism. (Yahoo Finance; CNN.)
Mar 9, 2026
LAWSUIT FILED: Anthropic sues the Department of Defense over the supply chain risk designation. (Axios.)
Mar 26, 2026
INJUNCTION: Judge Rita F. Lin (N.D. Cal.) grants a preliminary injunction blocking the § 3252 designation, the presidential directive, and the Hegseth directive, finding Anthropic likely to succeed on First Amendment, Fifth Amendment, and APA claims. (FindLaw; Mayer Brown.)
May 1, 2026
The Pentagon has struck AI deals with several big tech companies, and Anthropic and the Pentagon are reported back at the negotiating table. (CNN.)
Jun 30, 2026
Colorado AI Act rescheduled effective date. State preemption enforcement expected to begin.