The Pentagon is integrating Elon Musk’s Grok AI into military operations, handing the defense establishment a powerful tool trained on millions of Americans' social media posts while the chatbot faces active lawsuits alleging it generated child pornography from images of real victims.
The Department of Defense announced the launch this week, claiming Grok will let the military “execute missions faster and with greater precision” in areas from market research to supply chain management, according to the Washington Examiner. Over half of the Pentagon’s 3 million personnel already use the military's GenAI.mil platform. The Pentagon secured the initiative last year alongside contracts with Anthropic, Google, and OpenAI, each receiving $200 million.
The Examiner framed the story around Grok’s technical promise and Musk’s political standing, noting the AI is “viewed as one of the most promising AI tools due to its training on X’s data.” What the Examiner buried: Grok is also one of the most controversial models on the market. Last year, a code tweak designed to let the chatbot make “politically incorrect” statements went sideways. Within a day, the bot was praising Adolf Hitler, endorsing a new Holocaust, and threatening to sexually assault a user. The Examiner noted adjustments have prevented a repeat “of equal severity,” but acknowledged “lasting doubts about its capabilities.”
Those doubts are now playing out in federal court. The Norfolk Virginian-Pilot reports that a woman repeatedly raped as a child, whose abuse was photographed and circulated online, is suing xAI in San Jose U.S. District Court. The plaintiff, identified as “Jane Doe,” alleges Grok generated new child pornography using images of her and thousands of other victims.
The lawsuit claims xAI, aiming to add users, designed Grok “to respond to prompts to create and distribute sexual content using an identifiable person’s real image or video.” The UK-based Center for Countering Digital Hate reported Grok generated 3 million sexualized images in the 11 days after an image-editing feature went live last December, including manipulated images of more than 20,000 apparent children. Musk responded to the outrage by reposting an AI-generated image of a toaster in a bikini, saying he “couldn’t stop laughing.”
Doe’s lawsuit, overseen by Judge Nathanael Cousins, alleges “xAI and its founder Elon Musk chose to capitalize and profit from predators’ appetite for non-consensual sexual images and videos of real people, including children.” The suit seeks class-action status and a court order barring xAI from generating, possessing, or sharing child pornography.
Grok’s legal troubles don’t stop there. The Pilot reports xAI is fighting another lawsuit from four plaintiffs who say Grok created sexualized deepfake images of them. In a separate case, xAI is challenging Minnesota’s ban on apps used to digitally undress people, arguing the law is overly broad. xAI also sued a Texas man arrested for allegedly using Grok to create child sexual abuse material.
The Examiner noted Musk’s “usually positive relationship with President Donald Trump” helped his company secure the Pentagon deal. The revolving door between Big Tech and the defense establishment just spun a little faster. The question now is who controls the algorithms of war—and what happens when a tool the military trusts to manage supply chains is the same one courts say was used to exploit children.
