Pentagon Signs AI Deals But Leaves Out Anthropic
The Pentagon has signed deals with seven AI companies. They are SpaceX, OpenAI, Google, Nvidia, Reflection, Microsoft and Amazon Web Services.
The deals let these companies put their AI tools on secure US military networks. But AI company Anthropic was left out.
Anthropic and the Pentagon have been in dispute over possible AI misuse. Reports say the dispute is about Anthropic's refusal to let its AI be used for self-operating weapons or to watch Americans.
The Pentagon reportedly asked for unrestricted use of any AI it deploys. The Pentagon labeled Anthropic a supply-chain risk.
That label was previously used only for companies linked to foreign enemies. The Pentagon told staff to remove Anthropic tools within six months.
But a judge blocked that order. Anthropic had been the only AI company on the Pentagon's secure network.
Sources say many Pentagon staff think Anthropic's tools are better than the others. President Trump said last week that Anthropic was doing better.
This hinted at a possible change. Over 1.3 million Defense Department staff have already used the Pentagon's AI platform.
Key Terms 4
- Anthropic A company that builds AI tools and software
- supply-chain risk A label meaning a company is seen as a security threat
- injunction A court order that stops something from happening
- classified networks Highly secure computer systems used for secret military information