Microsoft Amicus Brief Supporting Anthropic's Legal Challenge

The Brief

On March 11, 2026, Microsoft Corporation filed an amicus curiae brief in the Northern District of California supporting Anthropic’s motion for a temporary restraining order against the Department of Defense’s supply chain risk designation. The brief represents an extraordinary intervention: a major Pentagon contractor publicly opposing a DoD procurement action on behalf of a competitor.

Microsoft’s brief argues that the designation “threatens to establish a precedent under which any technology company that publicly articulates safety boundaries for its products risks exclusion from the federal marketplace.” The brief contends this would create a perverse incentive structure in which AI companies that remain silent about safety limitations — or that privately negotiate constraints without public disclosure — receive preferential treatment over companies that communicate transparently with the public about how their technology should and should not be used.

Industry Chilling Effect

The core of Microsoft’s argument centers on the broader industry implications. Microsoft notes that it has its own published responsible AI principles and has publicly declined certain military applications of its technology in the past. If the government can designate a company as a supply chain risk based on the content of its published safety policies, Microsoft argues, then every AI company’s responsible use documentation becomes a potential liability rather than a public benefit. The logical corporate response would be to stop publishing safety commitments entirely — precisely the opposite of what policymakers, civil society, and the public have demanded from the AI industry.

Strategic Context

The filing is notable for what it does not say. Microsoft is OpenAI’s largest investor and strategic partner. OpenAI is the direct beneficiary of Anthropic’s exclusion from Pentagon contracts, having signed a $200 million agreement on the same day the Anthropic designation was issued. Microsoft’s decision to publicly support Anthropic against an action that commercially benefits its own portfolio company signals that the designation’s precedent is perceived as threatening enough to override competitive interests. Industry observers noted that Microsoft’s Azure Government cloud hosts both OpenAI and Anthropic models for federal customers, giving Microsoft a commercial stake in both companies’ government access.

Request to the Court

Microsoft’s brief urges the court to grant Anthropic’s TRO motion, arguing that the balance of harms overwhelmingly favors a temporary pause on implementation while the constitutional questions are resolved. The brief notes that the designation can be reimposed if the government ultimately prevails on the merits, but the competitive and reputational damage to Anthropic — and the chilling effect on industry-wide safety communications — cannot be undone after the fact.