The Pentagon has signed an agreement with Elon Musk’s xAI that could allow the Grok chatbot tool access to classified systems, according to a report.
The Defense Department agreed to a deal with Musk’s artificial intelligence company and is also hoping to reach a similar agreement with Google for its Gemini model, sources told The New York Times.
Officials are hoping to use those agreements to pressure AI giant Anthropic, a top Pentagon contractor, to lift its safeguards and allow the military greater access to its AI tool, Claude.
At the moment, Claude is the only AI model available in the military’s classified systems and is considered the best equipped to handle sensitive defense and intelligence work. The new deal means that Grok—which has generated non-consensual sexual images of women and children on X, as well as spouted racist rhetoric and praise of Adolf Hitler—could soon replace Claude.
Defense Secretary Pete Hegseth will hold crunch talks with Anthropic CEO Dario Amodei on Tuesday in what is believed to be a last-minute, desperate push to demand that the military be allowed to use Claude without its current restrictions, as long as its applications are lawful.
“Anthropic knows this is not a get-to-know-you meeting,” a senior Defense official told Axios. “This is not a friendly meeting. This is a s--t-or-get-off-the-pot meeting.”
Anthropic said it is willing to lift some of its restrictions but insists that guardrails remain in place to prevent the mass surveillance of Americans or the development of weapons that fire without human involvement.
Pentagon officials hope the pending deals will give them leverage, even though Musk’s xAI model is not considered as advanced or reliable as Anthropic’s Claude.
Grok has also frequently embarrassed Musk on X, the social media platform he owns.
X’s offices in Paris were raided by French prosecutors investigating the company after Grok was allegedly used to generate sexual images of women and children when prompted to do so.
Grok had already caused outrage after referring to itself as “MechaHitler,” last year, expressed skepticism about the murder of six million Jews during the Holocaust, and ranted about claims of “white genocide” in South Africa under unrelated posts.
Elsewhere, the White House is reportedly trying to block a bill in Florida that would introduce significant limitations on AI use, including restrictions on generating explicit images involving minors.
The White House has contacted Florida House Speaker Daniel Perez and urged him to oppose Gov. Ron DeSantis’ Artificial Intelligence Bill of Rights, sources told The Daily Signal. Officials are said to prefer a unified federal response to AI regulation rather than differing state-level laws.
A White House official told The Daily Signal that suggesting the administration’s opposition was motivated by a desire to avoid protections for minors was “fatally flawed.”
The Daily Beast has contacted the Pentagon and xAI for comment.
