Vcg | Visual China Group | Getty Images
Pentagon AI chief Cameron Stanley confirmed to CNBC that the Defense Department is expanding the use of Google’s Gemini artificial intelligence model, nearly two months after the DOD dropped Anthropic, citing it as a supply chain risk.
The DOD is using Google’s latest model for classified projects, according to a person with knowledge of the matter who asked not to be identified because details of the program are not public. The Information previously reported Google had signed an agreement with the DOD for a classified operation, citing a person familiar with the matter.
In addition to Gemini, the Pentagon is also working with OpenAI and other vendors to improve warfighting capabilities, Stanley told CNBC in a video interview.
“Overreliance on one vendor is never a good thing,” he said. “We see that, especially in software.”
The DOD’s acceptance of Google comes amid a heated legal dispute with Anthropic. Earlier this month, an appeals court in Washington, DC, denied Anthropic’s request to temporarily block the department’s delisting of the AI company as a counter-sanctions case.
That ruling came after a San Francisco judge, in a separate but related case, granted Anthropic a preliminary injunction preventing the Trump administration from enforcing a ban on the use of its Claude model. In separate rulings by two courts, Anthropic is excluded from DOD contracts but is able to continue working with other government agencies during the litigation.
A DOD spokesperson confirmed via email that the agency is not working with Anthropic at this time. President Donald Trump told CNBC last week that “it’s possible” there will be an agreement allowing Anthropic models to be used within the DOD.
Stanley said that by using Gemini, the Pentagon and US warfighters save time and money.
“There are many different things that save thousands of man-hours, thousands of man-hours every week,” he said.
A Google spokesperson said in an emailed statement that the company is part of a “broad organization” of companies that provide services and infrastructure “to support national security.”
“We support government agencies on both classified and non-classified projects, using our expertise in areas such as logistics, cybersecurity, diplomatic translation, vehicle maintenance, and critical infrastructure protection,” the spokesperson said.
The plan is facing opposition from within Google, with more than 700 employees signing a letter sent to Google CEO Sundar Pichai this week, demanding the company reject the split workloads. They said in the letter that they don’t want the technology to be “used in inhumane or more dangerous ways.”
The ultimate goal, according to Stanley, is to get the best possible outcome for America’s warfighters. To get there, the Pentagon must make sure it’s using AI models well.
“I have a quote that I usually say at these times, that you don’t cook a Thanksgiving turkey in the microwave,” he said. “You need to have the right technology to use it in the right way to achieve the right result.”
Stanley said the release of Anthropic Mythos earlier this month was a wake-up call. The powerful model has been made available to a limited number of companies, due in part to its advanced online capabilities and potential risks.
Stanley said the DOD “takes this very seriously” to “make sure we’re not only up to date but ready for what’s next, which is the full range of AI-enabled capabilities” in challenging environments.
-CNBC’s Jennifer Elias contributed to this report.
WATCH: Google, Pentagon in talks to use Gemini for decentralized systems
