Trump Memo Requires Pentagon to Diversify AI Vendors After Anthropic Dispute
President Trump signed a memo Friday requiring U.S. national security agencies to work with multiple artificial intelligence providers, a direct response to months of failed negotiations between the Pentagon and Anthropic over the company's safety restrictions.
The directive addresses a core legal issue: whether an AI vendor can impose conditions on how the military uses its tools. Anthropic demanded restrictions barring combat use without human oversight and blocking AI-powered surveillance. Pentagon officials rejected these conditions and designated Anthropic a supply-chain risk in late February.
The memo stops short of formally ordering agencies to abandon Anthropic but makes clear the government will not tolerate vendors that "subvert the chain-of-command" or threaten operational continuity. Agencies must end business with companies-including subcontractors-that disable or modify their technology without government approval.
What the Memo Actually Requires
Agencies including the Defense Department, CIA, and Treasury must seek assurances that AI systems cannot be disabled or severed during conflict. They must also avoid vendors that use AI to censor speech, conduct unlawful surveillance, or embed ideological bias.
The language acknowledges both sides of the Anthropic dispute. It protects the Pentagon's concern that a vendor could veto military workflows. It also addresses Anthropic's concern that the government could use AI in ways that violate constitutional rights.
Defense Secretary Pete Hegseth must update Pentagon policy on autonomous weapons systems to account for AI's evolving capabilities-Anthropic's other main demand.
Anthropic's Legal Position
Anthropic is challenging the Pentagon's supply-chain designation in court. The company won a preliminary injunction in California but failed to secure a stay in a parallel Washington case.
Meanwhile, the National Security Agency continues using Anthropic's Mythos product to find software vulnerabilities, including in Microsoft products. The Pentagon is winding down other Anthropic tools for military use.
Broader Policy Implications
The memo replaces Biden-era guidance on AI and national security. Trump administration officials view it as providing political cover for the Pentagon to quietly back away from its supply-chain risk claim while still diversifying vendors.
Trump told reporters Friday he is open to the U.S. government taking equity stakes in major AI labs and plans to discuss the idea with company executives next week.
For legal teams managing government contracts, the memo signals that vendor restrictions on AI use-particularly around military or national security applications-will face resistance. Agencies now have explicit direction to prioritize operational control and vendor diversity over any single provider's ethical guardrails.
Related reading: AI for Government covers public sector AI policy and procurement strategy. AI for Legal addresses compliance, contract analysis, and regulatory frameworks relevant to vendor agreements.
Your membership also unlocks:
