California Attorney General Rob Bonta announced an investigation into xAI, focusing on the proliferation of nonconsensual sexually explicit material generated by its Grok AI model. The announcement was made in Oakland.
Reports indicate that xAI may be facilitating the large-scale production of deepfake nonconsensual intimate images, which have been used to harass individuals, including women and girls, across the internet, including via the social media platform X.
Investigation Details
Attorney General Bonta stated that the investigation will determine whether and how xAI violated the law. He emphasized the state's commitment to ensuring AI safety and protecting California residents.
Over recent weeks, news reports have detailed instances of Grok users allegedly taking ordinary images of women and children and using Grok to depict them in suggestive and sexually explicit scenarios, or to "undress" them, without consent. xAI developed Grok's image generation models to include a feature described by the company as a "spicy mode," which generates explicit content. The company has reportedly used this mode as a marketing point.
Grok-generated images have reportedly been used to harass public figures and social media users. Reports have also described the use of Grok to alter images of children to depict them in minimal clothing and sexual situations, and to produce photorealistic images of children engaged in sexual activity. One analysis indicated that more than half of 20,000 images generated by xAI between Christmas and New Year depicted individuals in minimal clothing, with some appearing to be children.
Broader Context and Previous Actions
Attorney General Bonta has been involved in AI safety discussions, particularly concerning children:
- He has supported state legislation aimed at protecting children from AI companion chatbots.
- In September, Attorney General Bonta and Delaware Attorney General Jennings met with OpenAI to express concerns about its products' interactions with young people.
- In August, Attorney General Bonta sent a letter to 12 AI companies after reports of sexually inappropriate interactions between AI chatbots and children, reminding companies of their legal obligations to children as consumers.
- In 2023, Attorney General Bonta joined a bipartisan coalition of 54 states and territories in advocating for the creation of an expert commission to study AI's potential role in exploiting children through child sexual abuse material (CSAM).
- Attorney General Bonta has also opposed proposals for AI regulation bans and supports states' rights to implement AI safety measures.