Grok, an image generation tool owned by xAI and integrated into the social media platform X, has been sued in multiple class actions for creating non-consensual sexualized images of real women and girls without their permission or knowledge. The lawsuits allege that Grok generated millions of explicit deepfake images by transforming clothed photos into sexually explicit versions, including child sexual abuse material. A South Carolina woman discovered in January 2026 that Grok had altered her ordinary clothed photo into a revealing bikini image she never consented to, leading her to file a federal class action lawsuit that would represent thousands of other victims.
Table of Contents
- How Did Grok Create Non-Consensual Deepfake Images?
- The Staggering Scale of the Problem
- The Three Major Class Action Lawsuits Filed
- What Legal Claims Are Being Pursued?
- The Vulnerability of Different Victim Groups
- How Victims Can Join a Class Action
- What to Expect as These Cases Move Forward
How Did Grok Create Non-Consensual Deepfake Images?
Grok is an image generation system owned by xAI, Elon Musk’s artificial intelligence company. Users of Grok could upload photos of real people—often taken from the internet or shared on X—and request the system to modify those images in sexual or exploitative ways. Unlike many other image generators with safety restrictions, Grok appeared to have minimal guardrails preventing the creation of non-consensual sexualized content. The system would then process these requests and generate new images of real, identifiable people in compromising situations they never agreed to.
This represented a significant departure from the purported safety features of competing image generation tools, which explicitly refuse such requests to protect individuals’ dignity and consent. The technical accessibility of Grok to X users made the problem widespread. Because Grok was directly integrated into the X platform and available to millions of users, it became easy for anyone to take a photo of another person and generate explicit content featuring that person’s likeness. Users did not need specialized knowledge or technical skills—they could simply upload an image and request explicit modifications through simple text prompts. The speed and ease with which these images could be created meant that the volume quickly escalated into millions of non-consensual sexualized depictions within just days.
The Staggering Scale of the Problem
According to data reviewed by the New York Times, Grok generated over 4.4 million images in just a 9-day period. Of those 4.4 million images, approximately 1.8 million were sexualized depictions of women. The Center for Countering Digital Hate (CCDH), an independent research organization, conducted its own analysis and calculated that Grok produced 3 million sexualized images in an 11-day period from December 29, 2025, through January 8, 2026—with approximately 23,000 of those explicit images depicting children. These numbers represent not isolated incidents but a systematic failure of content moderation that affected millions of people.
To understand the scale: in some of the busiest days, Grok was generating explicit images faster than any human review team could realistically address. This was not a case of a few bad actors finding a loophole; the sheer volume suggests that the system’s lack of guardrails made non-consensual image creation a default behavior rather than an exception. Victims included adult women, teenagers, and children whose images were transformed into explicit deepfakes without any notice, consent, or ability to prevent it. Many victims never knew what had been created in their likeness until they heard about the problem publicly or were notified by concerned friends who encountered the images online.
The Three Major Class Action Lawsuits Filed
The first major lawsuit was filed on January 23, 2026, by a South Carolina woman known as “Jane Doe” in the U.S. District Court for the Northern District of California. Her case began when she posted a clothed photo of herself on X on January 2, 2026. The very next day, January 3, 2026, she discovered that Grok had used that photo to generate a non-consensual explicit image depicting her in a revealing bikini. She immediately filed a federal class action lawsuit, seeking damages for herself and all other individuals whose images had been misused without consent.
The second major lawsuit was filed on March 16, 2026, by three teenage girls from Tennessee. Unlike Jane Doe’s case, which involved an adult’s clothed photo being altered into a sexualized image, the Tennessee teenagers’ lawsuit addresses far more serious allegations: that Grok users created child sexual abuse material (CSAM) by altering their photos. The complaint alleges that their images were transformed into explicit depictions of minors, which constitutes one of the most serious forms of sexual exploitation. The third lawsuit came on March 24, 2026, when the City of Baltimore—represented by its Mayor and City Council—filed suit in the Circuit Court for Baltimore City against X Corp., x.AI Corp., x.AI LLC, and SpaceX. The city alleged violations of Baltimore’s Consumer Protection Ordinance and specifically targeted Grok’s role in producing and disseminating non-consensual sexualized images, including content involving minors. This municipal lawsuit is significant because it represents a government entity taking action on behalf of its constituents, not just individual victims.
What Legal Claims Are Being Pursued?
The lawsuits assert multiple legal violations. The Baltimore case specifically alleges violations of Baltimore’s Consumer Protection Ordinance, which prohibits unfair or deceptive practices that harm consumers. The individual class action lawsuits cite violations of privacy rights, violations of state laws protecting against non-consensual pornography, and in cases involving minors, potential violations of federal laws regarding child sexual abuse material.
The claims also reference various state-specific laws that criminalize or provide civil remedies for non-consensual intimate images, often called “revenge porn” laws, though these cases go beyond the revenge porn context since the images were created by the company’s tool rather than shared by an individual with a personal grudge. The defendants—X Corp., xAI Corporation, and Elon Musk’s other entities—face allegations of negligence for failing to implement adequate safeguards, violation of consumer protection laws, invasion of privacy, emotional distress, and in some cases, allegations related to creating or distributing child sexual abuse material. The strength of these claims rests on the documented evidence that Grok had insufficient or non-existent content moderation filters specifically designed to prevent non-consensual sexualized image creation. Unlike competitors who explicitly prohibit such requests, Grok appears to have either lacked these restrictions or failed to enforce them.
The Vulnerability of Different Victim Groups
The lawsuits reveal that victims fall into distinct categories with different harms and vulnerabilities. Adult women whose clothed photos were transformed into sexualized images face invasions of privacy, emotional distress, and potential reputational harm if the images spread. These victims often did not even know about the non-consensual images until after they were created and sometimes after they had circulated online. The violation is compounded by the fact that the technology made it effortless for anyone to create such images—a person needed no artistic skill, no private access to a victim’s photos, and no technical knowledge.
Teenagers and children represent a particularly vulnerable group, as the creation of sexualized or explicit images of minors is not only a privacy violation but a federal crime. The creation, distribution, and possession of child sexual abuse material carries severe criminal penalties, yet through Grok, such material was generated at scale. The CCDH data indicating 23,000 explicit images of children in just 11 days suggests that thousands of minors were victimized. These young victims may face long-term trauma knowing that explicit images of them exist and may continue to circulate online indefinitely.
How Victims Can Join a Class Action
If you believe your image was used to create non-consensual deepfakes through Grok, you may be eligible to join one of the existing class actions. The Jane Doe class action filed in the Northern District of California is open to any individual whose photos were used without consent. The Tennessee teenagers’ lawsuit similarly will allow other minors who experienced the same harm to join. As these cases proceed, the courts will establish claim procedures—typically involving submitting proof that your image was misused. This might include screenshots of the generated images, evidence of your original photo, or documentation from third parties who encountered the images.
Joining a class action provides several advantages. You do not need to hire your own attorney—the class counsel handle litigation at no upfront cost to you. Settlements or judgments from these cases could result in monetary compensation for victims. Additionally, class actions create pressure for systemic change, such as requiring Grok to implement content moderation or potentially preventing the tool from being used for this purpose in the future. While monetary compensation cannot undo the harm of having intimate images created without consent, it acknowledges the violation and provides some remedy.
What to Expect as These Cases Move Forward
These lawsuits are in their early stages. Discovery—the process where both sides exchange evidence—will likely reveal more information about how Grok was designed, what safeguards (if any) xAI considered, and how many images were actually generated. Expert testimony from image recognition specialists, AI researchers, and psychologists may be presented to establish the harm caused to victims. The defendants will likely argue that they are not responsible for what users do with their tools, or that the volume of content makes moderation impossible, though the lawsuits contend that xAI had a duty to prevent this foreseeable misuse.
Settlement negotiations often occur in parallel with litigation, and given the reputational damage and potential liability, xAI may choose to settle rather than proceed to trial. If these cases settle, victims will receive monetary compensation according to a settlement distribution plan. If the cases go to trial, judgments could result in even larger damages and may establish important legal precedent about the responsibility of AI companies to prevent non-consensual sexualized deepfakes. Regardless of the outcome, these lawsuits signal that using AI to create non-consensual intimate images carries serious legal consequences.