A federal judge has temporarily blocked two simultaneous actions against Anthropic: the Pentagon’s “supply chain risk” designation and a presidential directive ordering federal agencies to cease using the company’s AI products. The injunction, issued by Judge Rita F. Lin in late March 2026, is the first significant judicial check on executive authority over domestic AI procurement.
The court’s language was pointed. “At bottom, Anthropic has shown that these broad punitive measures were likely unlawful and that it is suffering irreparable harm from them,” Judge Lin wrote, according to reporting confirmed by CIO.com and corroborated across multiple outlets including CBS News and NPR. Three things make that sentence worth reading twice. The court found the measures likely unlawful, not merely procedurally flawed. It found irreparable harm, the standard required to justify a preliminary injunction. And it used the word “broad”, signaling the court’s concern that the designation swept further than the law permits.
The “supply chain risk” designation matters beyond this case. It’s a mechanism historically applied to foreign adversaries, companies with ties to hostile governments whose technology poses national security risks. Using it against a domestic AI company is, according to reporting on the court’s analysis, a significant expansion of the designation’s intended scope. Whether that expansion was lawful is now a live legal question.
Federal contractors have been the immediate practical losers in this situation. CIO.com’s coverage framed the ruling as “buying time for contractors to assess AI supply chains”, an accurate description of what a preliminary injunction does in procurement contexts. Agencies and contractors reportedly operating against an active compliance deadline to remove Anthropic technology now have a pause in that obligation. The specific timeline details haven’t been independently confirmed, but the practical effect of the injunction is clear: no one has to rip out Anthropic products while the merits are argued.
What to watch: this is a preliminary injunction, not a final ruling. The court found Anthropic is likely to succeed on the merits, that’s the legal standard for a preliminary injunction, but “likely” is not “certain.” A full merits hearing will follow. Watch for the Trump administration to either appeal the injunction or pursue the underlying case aggressively. Watch also for the exact ruling date to be confirmed from court records; the KSAT affiliate’s coverage is dated March 26, which may reflect a pre-ruling hearing rather than the injunction itself. Either way, the case is active and the next procedural development could shift the landscape.
The deeper compliance implication runs beyond Anthropic specifically. Federal procurement law contains multiple mechanisms for restricting vendors, and this ruling suggests courts will scrutinize whether those mechanisms are being applied within their intended scope. A “supply chain risk” label designed for foreign adversaries carries legal weight precisely because of that specificity. Stretching it to domestic companies is a different legal argument, and Judge Lin’s preliminary finding suggests it may not hold. For compliance officers tracking AI vendor risk in federal contracting, this case is now a leading indicator of where executive authority over AI procurement ends.
Status notice: This is a preliminary injunction. The ruling is temporary and subject to further proceedings. Monitor case developments before making procurement decisions based solely on the current injunction.