Anthropic Fights On: Seeks Full Court Rehearing Against Pentagon's AI Blacklist

AI powerhouse Anthropic is pushing back hard against the U.S. government, formally asking the full U.S. Court of Appeals for the D.C. Circuit to rehear its appeal against the Pentagon decision to blacklist its AI models. This latest move on October 13, 2026, comes after a split three-judge panel upheld the Pentagon 'National Security Supply-Chain Risk' designation last month, intensifying a high-stakes battle over who controls cutting-edge artificial intelligence and how it's used in warfare and surveillance. The tech firm argues this blacklisting, which blocks defense contractors from using its AI, is unlawful retaliation for its outspoken stance on AI safety and its refusal to allow its powerful Claude models for mass surveillance or autonomous weapons. The dispute ignited in March 2026 when the Pentagon labeled Anthropic a supply-chain risk after contract talks broke down over the company's ethical 'red lines'. While a D.C. Circuit panel narrowly sided with the government on September 25, 2026, another federal judge in California, Judge Rita Lin, ruled in August 2026 that a parallel designation by the Pentagon was illegal First Amendment retaliation. This created a peculiar situation where Anthropic has a win in California but a loss in D.C.. The Pentagon has already started shifting its classified AI workload away from Anthropic, reportedly moving 90% by September, with rivals like OpenAI and xAI stepping in to fill the void. Anthropic petition for a 'rehearing en banc' now puts the ball in the D.C. Circuit's court, where all active judges will decide whether to revisit the previous ruling. This isn't just about one company's contracts; it's a pivotal moment defining the boundaries of government oversight on private AI innovation and the future of AI ethics in national security. The outcome could set precedents for how AI companies navigate the complex terrain between commercial development and military applications, impacting billions in potential revenue and shaping global AI governance.