Anthropic faces ongoing uncertainty over its Pentagon “supply chain risk” label, imposed in early March 2026 after the company refused to drop AI safety limits on Claude that blocked use in mass domestic surveillance or fully autonomous lethal weapons. A California federal judge granted a preliminary injunction in late March, temporarily lifting the designation and giving Anthropic leverage with customers, but the Pentagon appealed to the Ninth Circuit while a parallel § 4713 designation under the Federal Acquisition Supply Chain Security Act remains active pending D.C. Circuit review. Recent moves by the Department of Defense to expand classified AI contracts with rivals OpenAI, Google, and xAI underscore competitive pressure, while a narrow carve-out for Anthropic’s Mythos cybersecurity model suggests selective pragmatism. Traders are watching upcoming appeals rulings and any settlement talks, as resolution hinges on whether courts uphold or overturn the unprecedented domestic-company designation before the market’s deadline.