In short
The criticism cites estimates that Grok generated as much as 3 million sexualized photographs in a matter of days, together with hundreds depicting minors.
Baltimore is in search of civil penalties, restitution for affected residents, and court docket orders to halt the alleged conduct.
Authorized consultants say the case might hinge on whether or not courts view AI techniques as lively creators of dangerous content material slightly than passive instruments.
Baltimore’s lawsuit in opposition to Elon Musk’s xAI and its Grok chatbot may determine how far cities can go to control synthetic intelligence within the absence of federal regulation, in accordance with one skilled.
The Mayor and Metropolis Council of Baltimore have sued X Corp., xAI, and SpaceX in a Maryland court docket, alleging the businesses violated native client safety legal guidelines by designing and deploying Grok, a generative AI system accused of manufacturing and spreading non-consensual sexualized photographs, together with of minors.
The lawsuit claims Grok allows customers to “undress” or “manipulate photographs” of actual folks with minimal prompting, exposing residents to privateness violations and psychological hurt, in accordance with a assertion from regulation agency DiCello Levitt, which is representing town alongside the Baltimore Metropolis Legislation Division.
“These deepfakes, particularly these depicting minors, have traumatic, lifelong penalties for victims,” Baltimore Mayor Brandon M. Scott stated within the assertion.
The case arrives amid rising international scrutiny of Grok, together with investigations throughout the U.S., EU, France, UK, Australia, and Eire, in addition to a federal class motion filed final week by three Tennessee minors alleging the instrument generated youngster sexual abuse materials utilizing their actual photographs.
“This lawsuit might be seen as a strategic transfer by a metropolis to control AI within the absence of federal laws, utilizing client safety and public hurt doctrines to convey AI firms inside its enforcement ambit,” Ishita Sharma, managing accomplice at Fathom Authorized, informed Decrypt.
“On legal responsibility, whereas customers prompting dangerous content material will likely be a part of the argument, the stronger authorized focus is more likely to be on whether or not the AI system itself materially contributed,” Sharma stated, including that if courts view Grok “as an lively creator slightly than a passive middleman,” duty will fall extra closely on xAI.
By design
The Baltimore swimsuit alleges the businesses “designed, marketed, and deployed” Grok, figuring out it may generate non-consensual intimate imagery and materials resembling youngster sexual abuse content material, whereas publicly claiming such content material was prohibited.
The criticism cites estimates that Grok generated between 1.8 million and three million sexualized photographs in simply days between December 29, 2025, and January 8, 2026, together with round 23,000 depicting kids, in accordance with the Middle for Countering Digital Hate and a New York Instances evaluation.
Baltimore alleges the surge was partly triggered after Elon Musk amplified Grok’s image-editing function by responding “Good” to a bikini picture of himself generated by the instrument, with output rising from roughly 300,000 photographs within the 9 days earlier than his put up to just about 600,000 per day on X.
“X is now one of many largest distributors of NCII and CSAM,” the lawsuit reads, citing the defendants’ personal platform insurance policies banning such content material as proof of misleading misrepresentation.
“Proof of delayed safeguards or inaction within the face of identified dangers would strengthen claims of negligence or recklessness,” Sharma stated, including that dismissal of the swimsuit seems unlikely, with settlement probably the most possible end result, although the case may nonetheless lead to “a precedent-setting ruling on AI accountability.”
The town is in search of civil penalties, injunctive reduction to halt the illegal conduct, restitution for affected residents, and disgorgement of ill-gotten earnings.
Decrypt has reached out to Musk through xAI and SpaceX for remark.
Day by day Debrief E-newsletter
Begin every single day with the highest information tales proper now, plus authentic options, a podcast, movies and extra.
