In a news release, the Defence Department said the contract award is worth up to US$200 million ($335m). The department issued similar awards to Google, Anthropic and OpenAI, it said.
“Today’s awards bring in the best US-based frontier AI talent to help apply cutting-edge AI to solve DoD use cases,” said the announcement, from the agency’s Chief Digital and Artificial Intelligence Office.
Grok came under fire last week after launching into an anti-Semitic rant and invoking Adolf Hitler after it was a programmed to be less politically correct.
The incident prompted the company to say it would improve its model.
A day later, xAI unveiled a sweeping update that it claimed put Grok on the cutting edge of AI development.
However, the damage was done. The incident last week demonstrated the pitfalls of rapid deployment of new technology in the AI arms race and the potential consequences of training flaws or the manipulation of existing models by users.
The Defence Department did not immediately respond to a request for comment beyond its news release.