The US military is set to significantly increase its utilization of artificial intelligence (AI) following the Pentagon's agreement to new and expanded contracts with several major technology companies. These eight agreements involve Google, OpenAI, Amazon, Microsoft, SpaceX, Oracle, Nvidia, and the startup Reflection. Under these new arrangements, the Pentagon stated that AI technology will now be employed for any "lawful operational use." The Pentagon explicitly said, "These agreements accelerate the transformation [of] the US military as an AI-first fighting force."
A notable absence from these agreements is Anthropic. The company has reportedly expressed concerns regarding the potential applications of its AI tools by the Pentagon, both in warfare and for domestic purposes. Anthropic is currently engaged in a lawsuit against the government, alleging that it faced retaliation after refusing to accept contract clauses that permitted "any lawful use" of its technology.
The Pentagon, in its statement on Friday, highlighted that its strategy of partnering with numerous companies on AI development is intended to help it avoid "vendor lock." This approach aims to prevent over-reliance on any single company for critical technology. The Pentagon stated, "Access to a diverse suite of AI capabilities from across the resilient American technology stack will give warfighters the tools they need to act with confidence and safeguard the nation against any threat."
According to the Pentagon, over a million individuals across the defense department have utilized the military's AI platform since its launch last year. This platform hosts various AI tools that have reportedly helped reduce the time required for numerous tasks, transforming processes that once took months into mere days. The Pentagon views access to powerful technology as a key component for success in modern warfare and has been actively working to enhance its AI capabilities for several years.
Anthropic's AI tools, including a version of its Claude chatbot, are reportedly still in use within many US government and defense agencies. This is partly because Anthropic was among the first AI companies to have its technology deployed for classified work. However, the relationship reportedly deteriorated earlier this year. Anthropic's chief executive, Dario Amodei, publicly voiced concerns that powerful AI tools could be misused by defense agencies for mass domestic surveillance or for the deployment of fully autonomous weapons. In response to these concerns, Defense Secretary Pete Hegseth reportedly designated Anthropic as a "supply chain risk," deeming its technology too dangerous for use in government settings. Anthropic's legal challenge to this designation is anticipated to go to court in September.
This dispute with Anthropic appears to have created an opportunity for other AI companies to deepen their collaboration with the government and military. OpenAI, the developer of the ChatGPT chatbot, was reportedly the first company to finalize a new contract with the Pentagon following the issues with Anthropic. The company signed its agreement at the end of February. An OpenAI spokeswoman confirmed that the Pentagon's announcement on Friday served as a formalization of that existing deal. "As we said when we first announced our agreement several months ago, we believe the people defending the United States should have the best tools in the world," the spokeswoman reportedly said.
While Google's Gemini chatbot was already in use by some government entities, this new agreement marks the first time the chatbot will be utilized for government work at a classified level. Earlier this week, the BBC reported that hundreds of Google employees, including many from its AI research division DeepMind, sent a letter to CEO Sundar Pichai urging the company not to deepen its collaboration with the government. A Google spokesperson reportedly did not reply to a request for comment.
SpaceX, through its AI startup xAI, is also included in the new agreements. Elon Musk's xAI operates the Grok chatbot. While xAI is widely considered to offer AI capabilities that are less advanced than some of its competitors, it contributes to the military's diverse AI portfolio. A representative for SpaceX reportedly did not respond to a request for comment.
Nvidia and the startup Reflection are set to have their open-source AI models, Nemotron and Reflection 70B respectively, utilized by the government. Nvidia has reportedly clarified that it is not supplying any hardware as part of this specific deal. Nvidia reportedly declined to comment, and Reflection reportedly did not respond to a request for comment.
Microsoft, Amazon Web Services (AWS), and Oracle have historically provided cloud services that are purpose-built to enable government operations conducted online. Microsoft and AWS reportedly did not reply to a request for comment. Oracle stated that its defense work "enables the Department of War to build, deploy, and scale any model, without vendor lock-in." Friday's announcement from the Pentagon signifies a continuation and expansion of these services, which will now be leveraged to deploy a greater number of AI models and tools for military applications than ever before.
