Elon Musk’s xAI inks Pentagon deal for contentious Grok chatbot

Grok is the AI chatbot available through the social media platform X. (Jonathan Raa/NurPhoto/Getty Images/File via CNN Newsource)

NEW YORK — Elon Musk’s xAI, which features a large language model that has spewed Hitler-supporting rhetoric and antisemitic tropes, said Monday it has signed a deal to provide its services to the U.S. Department of Defense.

Launched at the end of 2023, Grok has rarely been out of the headlines for its offensive gaffes, and will now offer its services as “Grok for Government.”

In addition to the Pentagon contract, “every federal government department, agency, or office (can now) purchase xAI products” thanks to its inclusion on an official supplier list, xAI added.

After an update on July 7, the chatbot praised Adolf Hitler in some responses, denounced on X “anti-white hate”, and described Jewish representation in Hollywood as “disproportionate.”

xAI apologized on Saturday for extremist and offensive messages, and said it had corrected the instructions that led to the incidents.

The new version of the chatbot, Grok 4, presented on Wednesday, consulted Musk’s positions on some questions it was asked before responding, an AFP correspondent saw.

The contract between xAI and the U.S. Department of Defense comes even as Musk and President Donald Trump are locked in a bitter feud.

The two men became close during Trump’s latest run for the presidency and, following the inauguration, the Republican billionaire entrusted Musk with managing the new agency known as DOGE to slash the government by firing tens of thousands of civil servants.

After ending his assignment in May, the South African-born entrepreneur publicly criticized Trump’s major budget bill for increasing government debt.

The president and the businessman engaged in heated exchanges on social media and in public statements before Musk apologized for some of his more combative messages.

The government and the defence sector are considered a potential growth driver for AI giants.

Meta has partnered with the start-up Anduril to develop virtual reality headsets for soldiers and law enforcement, while in June OpenAI secured a contract to provide AI services to the U.S. military.