The Pentagon has secured new AI agreements with Nvidia, Microsoft, Amazon Web Services and Reflection AI for use on classified military networks, expanding its options after its relationship with Anthropic collapsed. The deals mirror previous arrangements with SpaceX, OpenAI and Google, providing access to both proprietary and open-source AI models.

Anthropic's Claude was previously the sole AI model used in classified settings, including operations in the war against Iran and the capture of Venezuelan leader Nicolás Maduro. However, CEO Dario Amodei refused to sign a new deal without guarantees that Anthropic's tools would not be used for mass domestic surveillance or lethal autonomous weapons. Defence secretary Pete Hegseth called Amodei an "ideological lunatic" during Congressional testimony.

Pentagon chief technology officer Emil Michael told CNBC that open-source AI represents a "new vector" for the department because Chinese open-source models have infiltrated numerous companies. The US wants an American alternative. Meanwhile, more than 560 Google employees signed an open letter urging CEO Sundar Pichai to abandon the Pentagon deal over concerns about "inhumane" applications. These agreements position the US military to accelerate AI adoption across battlefield operations.