Pentagon AI chief confirms work with Google after Anthropic blacklist

Pentagon AI chief Cameron Stanley confirmed to CNBC that the Department of Defense is expanding its use of Google’s Gemini AI model, nearly two months after the Department of Defense dropped Anthropic, calling it a supply chain risk.

The Defense Department is using Google’s latest model for classified projects, according to a person with knowledge of the matter who asked not to be named because details of the arrangement have not been made public. Information before reported HE Google He had signed an agreement with the Defense Department for classified work, according to a person familiar with the matter.

In addition to Gemini, the Pentagon is also working with OpenAI and other providers to modernize wartime capabilities, Stanley told CNBC in a video interview.

“Over-confidence in a vendor is never a good thing,” he said. “We see this especially in software.”

The Department of Defense’s adoption of Google comes amid a heated legal dispute with Anthropic. Earlier this month, a federal appeals court in Washington, D.C., rejected Anthropic’s request to temporarily block the department from blacklisting the AI ​​company while a lawsuit challenging the sanction continues.

The ruling comes after a judge in San Francisco, in a separate but related case, granted Anthropic an injunction preventing the Trump administration from banning the use of the Claude model. Due to split decisions from two courts, Anthropic was excluded from Department of Defense contracts but can continue to work with other government agencies during the litigation.

A Department of Defense spokesperson confirmed via email that the agency is not currently working with Anthropic. President Donald Trump told CNBC last week that a deal allowing Anthropic’s models to be used at the Department of Defense was “possible.”

By using Gemini, the Pentagon and U.S. warplanes are saving time and money, Stanley said.

“There are a lot of different things that save thousands of man hours, literally thousands of man hours on a weekly basis,” he said.

The regulation faces some opposition within Google; More than 700 employees signed a letter sent to Google CEO Sundar Pichai this week calling on the company to reject confidential workloads. They said in the letter that they did not want the technology to be “used in inhumane or extremely harmful ways.”

According to Stanley, the overall goal is to achieve the best outcome for American warfighters. To get there, the Pentagon needs to make sure it’s using AI models correctly.

“I have a personal saying that I usually say in these moments, you don’t microwave a Thanksgiving turkey,” he said. “To achieve the right outcome, you need to have the right technology for the right use case.”

Stanley said Anthropic’s launch of Mythos earlier this month was a wake-up call. The powerful model was made available to a limited number of companies due to its advanced cyber capabilities and the potential risks it poses.

Stanley said the Department of Defense is “taking this very seriously” so that “we can ensure that we’re not only seizing the moment, but also being prepared for what’s next, which is a set of AI-enabled capabilities in challenging domains.”

—CNBC’s Jennifer Elias contributed to this report.

WRISTWATCH: Google and Pentagon in talks to embed Gemini in classified systems