Pentagon AI chief confirms DOD's use of Google, says reliance on one model 'never a good thing'

Cameron Stanley, the Pentagon's AI lead, has revealed that the Department of Defense (DOD) is increasing its reliance on Google's Gemini artificial intelligence model. This announcement comes just two months after the DOD severed ties with Anthropic, citing concerns over supply chain risks. According to sources familiar with the situation, the DOD is utilizing Google's advanced AI for classified initiatives, although specific details remain confidential. Reports indicate that Google has formalized an agreement with the DOD for these classified projects. In addition to Gemini, the Pentagon is also collaborating with OpenAI and other tech providers to enhance its military capabilities. Stanley emphasized the importance of diversifying partnerships, stating, "Overreliance on one vendor is never a good thing," particularly in the realm of software development. This shift towards Google occurs against the backdrop of ongoing legal challenges with Anthropic. Recently, a federal appeals court in Washington, D.C., declined Anthropic's request to lift the DOD's ban on its services while litigation continues. Meanwhile, a separate ruling in San Francisco allowed Anthropic to maintain its operations with other government entities outside the DOD. A DOD representative confirmed via email that there are currently no active contracts with Anthropic. In a recent interview, former President Donald Trump hinted at the possibility of a future agreement that would allow Anthropic's models to be used by the DOD. Stanley highlighted the efficiency gains from using Gemini, noting that it is significantly reducing the time and resources needed for military operations. "We're saving thousands of man hours on a weekly basis," he explained. However, this partnership has faced internal pushback at Google, where over 700 employees signed a letter urging CEO Sundar Pichai to reject classified projects. These employees expressed concerns about the potential misuse of technology in harmful ways. Ultimately, Stanley's goal is to ensure that the Pentagon effectively leverages AI technology to support U.S. warfighters. He remarked, "You don’t cook a Thanksgiving turkey in the microwave; you need the right technology for the right use case to achieve the right outcome." He also referred to Anthropic's recent launch of its Mythos model as a critical moment for the DOD, emphasizing the need for caution due to its advanced capabilities and associated risks. The Pentagon is committed to preparing for future challenges and ensuring responsible use of AI technologies.