SpaceXAI, Stability AI sued over AI-generated CSAM deepfakes
What's the story
Elon Musk's artificial intelligence (AI) firm, SpaceXAI, and Stability AI have been slapped with a class-action lawsuit. The plaintiffs allege that these companies' image-generation models were misused to create child sexual abuse material (CSAM). The lawsuit also claims that SpaceXAI did not provide enough information about a suspected perpetrator to law enforcement agencies.
Case expansion
Stability AI added as a defendant in the case
Two new plaintiffs, one from Wyoming and another from Wisconsin, have joined the original lawsuit filed by three Tennessee teenagers. All five plaintiffs are identified as Jane Does in court documents. The amended complaint also adds Stability AI as a defendant in the case. The lawsuit alleges that perpetrators used these companies' AI models to modify photos of underage individuals into CSAM.
Allegations detailed
Plaintiffs seeking monetary compensation for their suffering
The plaintiffs have accused SpaceXAI and Stability AI of creating CSAM, profiting from sex trafficking operations, negligence, defective product design, and public nuisance. Annika Martin, an attorney for the plaintiffs, said they want to impose "guardrails" on these companies to prevent further harm to future generations. The plaintiffs are seeking stronger measures from these AI firms against the creation of exploitative imagery and monetary compensation for their suffering.
Reporting concerns
SpaceXAI accused of failing to report suspected child sexual exploitation
The lawsuit also highlights SpaceXAI's alleged failure to report suspected child sexual exploitation, including CSAM, to the National Center for Missing and Exploited Children (NCMEC). This comes despite the fact that internet companies are legally required to do so. The complaint details how Jane Doe 4's stepfather used SpaceXAI's chatbot, Grok, to create thousands of sexually explicit images from a single photo of her when she was about 11 years old.
Reporting issues
Allegations of non-compliance with legal requirements
The lawsuit claims SpaceXAI only tipped NCMEC about Jane Doe 4 when her stepfather requested an image showing her being raped by multiple men. The company didn't include any of the abusive images in its report and also failed to provide the alleged perpetrator's IP address despite repeated requests from law enforcement agencies. This has raised questions over SpaceXAI's compliance with legal requirements for reporting suspected child sexual exploitation cases.
AI involvement
Lawsuit implicates Stability AI over open-weight image-generating tools
The lawsuit also implicates Stability AI, the maker of the Stable Diffusion image generator. It claims that an app on a perpetrator's phone used to create AI CSAM of the plaintiffs relied on Stability AI's image-generating tools. Unlike proprietary models from companies like SpaceXAI and OpenAI, Stability AI's models are open-weight, making it easier for users to bypass restrictions put in place by model makers.