D.C. CIR.Research
SSRN: Thomas Lee argues courts should not give frontier AI firms constitutional claims in Pentagon disputes
In an October 13, 2026 SSRN essay, Thomas H. Lee argues that courts should not turn frontier AI companies' disputes with the federal government over military use of their models into constitutional cases. Responding to Anthropic's lawsuits over its supply-chain-risk designations, he contends a classified contractor's public campaign to control product use is not protected speech and that limits should come from Congress.
Key points
- Lee posted the essay on SSRN on October 13, 2026.SSRN
- It addresses Anthropic's suits after the Pentagon designated it a supply-chain risk for refusing to authorize all lawful uses of Claude.SSRN
- Lee recounts that a California district court ruled for Anthropic on August 27 and a divided D.C. Circuit panel ruled against it September 25.SSRN
- Even if speech were implicated, he says the Pickering-Umbehr framework weighs those interests against deference to presidential command of military operations.SSRN
- He calls the due process claim weaker, citing Anthropic's rising valuation and existing procedural protections in procurement statutes.SSRN
- He argues limits on presidential military uses of AI should come from Congress and the people, not court-made constitutional rights.SSRN