Tennessee Teens Sue Grok, Stability AI Over Explicit Deepfakes

Landmark Lawsuit Targets AI Image Generators

Two teenage girls from Tennessee have filed a federal lawsuit against xAI, the company behind the Grok chatbot, and Stability AI, the developer of the Stable Diffusion image generator. The plaintiffs allege that the companies' AI tools were used to produce sexually explicit deepfake images of them, in violation of Tennessee's Ensuring Likeness Voice and Image Security (ELVIS) Act. The lawsuit, filed in the U.S. District Court for the Middle District of Tennessee, is one of the first to directly challenge AI companies under a state law specifically designed to protect individuals from unauthorized digital replicas.

According to the complaint, the girls — identified only as Jane Doe #1 and Jane Doe #2 — discovered the altered images circulating online in early 2025. They say the images were created using both Grok's image generation feature and Stability AI's models, which allow users to generate photorealistic content from text prompts. The lawsuit accuses the companies of negligence, violation of the ELVIS Act, and infliction of emotional distress. The families are seeking unspecified damages and a court order requiring the companies to implement stricter safeguards against the creation of non-consensual intimate imagery (NCII).

The ELVIS Act: A New Legal Frontier

Tennessee's ELVIS Act, which took effect on July 1, 2024, was originally elevated as a measure to protect musicians' voices and likenesses from AI cloning. Named in honor of Elvis Presley, the legislation expands the state's right-of-publicity laws to cover artificial intelligence-generated replicas. It explicitly makes it unlawful to create or distribute a digital replica of a person's voice or likeness without consent, including in cases where the replica is "so realistic" that it could deceive a reasonable person. This marks the first time a U.S. state has enacted a law that directly addresses AI-generated fakes in such broad terms.

The Tennessee lawsuit argues that both xAI and Stability AI are liable because their products lack adequate filters to prevent users from generating explicit content involving real minors. The plaintiffs' attorney, Carrie A. Huff, said in a statement: "These companies have unleashed technology that can devastate a child's life, and they have done so without meaningful guardrails. We intend to prove that the ELVIS Act applies to them, and that they must be held accountable."

How the Deepfakes Were Created and Shared

The complaint details how the images were allegedly made. Using publicly available photos from the girls' social media profiles, someone fed the images into AI tools that can "swap" faces onto explicit body images or generate entirely new scenes. The resulting fakes were then shared on a popular messaging platform, where they were viewed and reposted multiple times. The lawsuit does not name the individual who created the images — likely because the identity remains unknown — but instead targets the toolmakers, arguing that their products are defective because they can be used for this purpose.

Legal experts note that this approach is a strategic shift. Rather than pursuing individual harassers, the families are going after the upstream providers, hoping that financial pressure will force faster adoption of safety measures. "The success of this case could hinge on whether courts view AI image generators as neutral tools or as products with a foreseeable risk of misuse," said Dr. Emily Landon, a professor of digital ethics at Vanderbilt University. "In product liability law, manufacturers are expected to design reasonably safe products. The question here is whether OpenAI's Grok and Stability AI's models meet that bar."

Grok and Stability AI Respond

xAI, founded by Elon Musk, has not issued a public statement specifically about the lawsuit. However, the company has previously defended Grok's image generation as a free-speech feature, and Musk has been a vocal critic of AI regulation. In internal communications reviewed by news agencies, xAI's safety team reportedly flagged concerns about non-consensual imagery in late 2024, but changes to the system were slow-rolled. The lawsuit alleges that Grok's image generator does not require users to verify they are adults, and it lacks a watermarking system that would make AI-generated images easier to identify.

Stability AI, a London-based startup, has responded more directly. In a statement released to press, the company said: "Stability AI takes abuse of our models extremely seriously. We prohibit the generation of explicit content involving real people, and we invest heavily in safety classifiers and takedown procedures. We are reviewing the lawsuit and will respond in due course." Stability AI has faced prior legal battles over copyright infringement from artists and stock photo agencies, but the Tennessee case is the first to center on personal likeness and NCII.

The Mental Health Toll and Wider Impact

For the victims, the lawsuit is as much about personal justice as it is about systemic change. The girls describe anxiety, depression, and social withdrawal in the complaint. One of them says she stopped attending school for two weeks after the images surfaced. "Every time a phone buzzes, I panic," she is quoted in the filing. "I don't know who has seen them, and I can't erase that from people's memories."

The case also arrives amid a broader national conversation about AI safety. In January 2025, the U.S. Federal Trade Commission announced an investigation into generative AI companies over deceptive practices. Congress, meanwhile, has repeatedly failed to pass a federal law against deepfakes, leaving a patchwork of state statutes. Tennessee's ELVIS Act is considered one of the strongest, but legal scholars warn that enforcing it across state lines is difficult. "If a user in California uses a server in Oregon to generate an image of a Tennessee resident, which court has jurisdiction? This case will test that question," noted Professor Gerald Okafor of the University of Tennessee College of Law.

Precedent and the Road Ahead

Legal observers say the Tennessee lawsuit could have ripple effects beyond xAI and Stability AI. Other major platforms — including OpenAI's DALL-E, Google's Gemini, and Meta's AI tools — are watching closely. If the plaintiffs succeed, AI companies may be required to implement age-verification, real-time filtering of known individuals, or even prohibition of photorealistic human generation altogether. Consumer advocacy groups have applauded the suit. The nonprofit Encode Justice, which campaigns for AI accountability, called it "a watershed moment for holding AI powerful firms responsible for the harm their products enable."

However, the case is not guaranteed to survive the early stages. The defendants will likely argue that the ELVIS Act was never intended to apply to toolmakers, only to direct creators of digital replicas. They may also invoke Section 230 of the Communications Decency Act, which shields online platforms from liability for user-generated content. Courts have applied Section 230 inconsistently to AI output, making the outcome far from certain.

What's Next

The next legal milestone will be a motion to dismiss, expected within 60 days. If the court allows the case to proceed, both companies will be required to hand over internal safety documents and model training data, which could reveal how and when they became aware of NCII risks. A trial, if not settled before then, is unlikely to begin before 2026.

For the two Tennessee families, the fight is just beginning. Their attorney says they are determined to see it through, not only for themselves but for the many other children who may be victims of AI-generated abuse. "Right now, there are hundreds of thousands of images like these being made every day," Huff said. "Our goal is to make the companies that enable this to feel the same fear and helplessness our clients feel."

A Call for Federal Action

The case also adds momentum to calls for a federal deepfake law. In February 2025, bipartisan senators reintroduced the No Fakes Act, which would create a federal right to control one's voice and likeness in AI-generated content. That bill remains stalled in committee, but high-profile litigation can often shake loose political action. "Historically, Congress responds to lawsuits that make headlines," Professor Okafor said. "If the Tennessee case results in a major verdict, you will see renewed urgency in Washington."

Until lawmakers catch up, tools like the ELVIS Act remain the primary shield for ordinary Americans. The Tennessee lawsuit is a test of whether that shield can stop a multi-billion-dollar industry. The next court hearing is scheduled for late spring, and both sides are expected to present expert testimony on the technical capabilities of the AI models and the adequacy of existing safeguards.