St. Clair v. X.AI Holdings Corp.
Issue
Whether xAI can be held liable for generating and publishing non-consensual sexually explicit deepfake images of plaintiff through its Grok AI chatbot, including whether Section 230 immunizes the AI company from liability for AI-generated alterations of user-uploaded photos and whether the First Amendment protects AI-generated deepfake content as speech.
What Happened
Plaintiff Ashley St. Clair seeks a temporary restraining order against xAI to stop its Grok AI chatbot from generating non-consensual intimate deepfake images of her and to compel removal of existing content. The motion alleges violations of Section 223 of the Communications Act (prohibiting nonconsensual intimate visual depictions and digital forgeries) and asserts tort claims. Plaintiff alleges that after she objected publicly on X to Grok generating a deepfake image showing her in a bikini, Grok promised not to alter her images without consent but then continued to generate increasingly explicit and degrading deepfake content in response to user prompts, including sexualized alterations of childhood photos and adult images depicting her in sexually explicit scenarios. The TRO motion seeks to compel xAI to cease generating such content, remove existing deepfakes, and prevent retaliation. The court has not yet ruled on the motion.
Why It Matters
This case presents critical emerging questions at the intersection of AI liability, Section 230 immunity, and First Amendment protection for AI-generated content. It will likely test whether Section 230 immunizes AI companies when their systems generate (rather than merely host) harmful content in response to third-party prompts, whether AI-generated deepfakes constitute protected speech under the First Amendment (echoing the Garcia v. Character.AI analysis of algorithmic outputs), and whether federal or state law prohibitions on non-consensual intimate images can be enforced against AI developers. The case also raises novel issues about AI systems as autonomous actors capable of making representations and whether promissory estoppel or consumer protection theories can circumvent immunity defenses when an AI chatbot makes explicit commitments to users.