Baltimore Sues xAI Over Grok Sexual Deepfakes

The city of Baltimore sued Elon Musk’s xAI on March 24.

The complaint, filed in Baltimore Circuit Court, accuses the company of violating the city’s consumer protection ordinance by marketing Grok as a safe, general-purpose artificial intelligence (AI) assistant while the chatbot generates and distributes nonconsensual sexually explicit deepfakes.

Lawyers for the city of roughly 568,000 residents say the tool has become one of the largest distributors of such material, including images depicting children, despite explicit promises to ban it.

“Unlike traditional consumer software tools, Grok is designed to generate novel content in response to user prompts, drawing on vast quantities of training data that include scraped online material, images, and text,” the city said in the lawsuit.

Referring to X and the stand-alone Grok app, the complaint said, “Both channels permit users to ask Grok to ‘undress’ or ‘nudify’ photos of third parties—ranging from celebrities to private citizens, including children—placing them in sexually suggestive, degrading, or violent scenarios.”

The complaint cites an independent analysis showing that Grok generated an estimated 3 million realistic-looking sexualized images—more than 23,000 of them apparently involving children—in just 11 days near the start of the year.

The lawsuit seeks civil penalties, an injunction forcing changes to Grok’s design and features, restitution, and disgorgement of profits.

xAI, based in the San Francisco Bay Area, developed Grok as a generative AI tool marketed as a free-speech AI model and creative tool.

xAI did not immediately return The Epoch Times’ request for comment.

The Baltimore filing is the latest lawsuit xAI has faced amid similar complaints about Grok’s image-generation tools.

Three people from Tennessee, including two minors, sued xAI on March 16. They alleged the chatbot generated sexually explicit images by digitally altering their real photographs. The Tennessee plaintiffs, who filed in federal court in California and seek class-action status, claimed that the AI tool created and helped spread the harmful content.

Similar to the Baltimore case, the Tennessee suit alleges that xAI designed Grok without adequate safeguards, allowing users to easily create nonconsensual deepfakes.

Regulators in multiple countries have opened probes into Grok because of similar concerns about child safety and nonconsensual content. In mid-January, xAI restricted some image-editing features after public outcry.

The company has long promoted the tool as helpful for everyday users while claiming that it has strict rules against nonconsensual sexual content and child exploitation material.

Musk said in a January X post that he was “not aware of any naked underage images generated by Grok.”