The Pentagon has signed on to use Grok, the AI chatbot built by Elon Musk’s company xAI, as part of a new $200 million agreement that opens the door for its deployment across the federal government, the company announced.

The announcement comes during Musk’s public ‘breakup’ with President Trump and days after Grok generated antisemitic responses and praised Adolf Hitler.

The rollout is part of “Grok for Government,” a newly launched suite of tools designed for use by federal agencies, local governments, and national security operations. xAI said its products, including its latest Grok 4 model, will now be available for purchase through the General Services Administration (GSA), allowing any federal office to adopt the technology.

The adoption of Grok aligns with the Trump administration’s push for more aggressive utilization of artificial intelligence across the government. Since taking office, Mr. Trump has been a proponent AI as a mainstay of national security and innovation.

Musk briefly served in the Trump administration earlier this year, overseeing the White House’s Department of Government Efficiency, or DOGE, before stepping down in May amid a public break with Mr. Trump over his sweeping tax and spending bill. Musk has since emerged as an imposing critic of that legislation, even floating the idea of launching a third political party.

The announcement comes just days after Grok generated antisemitic responses to user prompts and referenced Hitler as part of what the company called an effort to make the model “less politically correct.” Hours later, Musk wrote in a post on X that “Grok was too compliant to user prompts. Too eager to please and be manipulated, essentially. That is being addressed.”

The posts were later deleted and xAI said it “quickly” patched the issue. One day later, xAI launched an upgraded version of Grok it described as a major leap forward. Musk also announced that Grok would be used in Teslas.