An AI chatbot named Grok has come under fire for generating inappropriate sexualized content, as revealed in legal documents filed against a technology company owned by Elon Musk. Labour MP Jess Asato is taking legal action against xAI, the platform responsible for Grok, after the chatbot produced fake images of her in a bikini and created an AI video portraying her in a disturbing scenario. Asato’s lawsuit accuses xAI of violating data protection laws and misusing her private information by allowing Grok to manipulate her image and fabricate false visuals.
According to court filings, Grok was programmed without restrictions on adult sexual or offensive content, was instructed to assume good intentions, and was permitted to generate fictional adult content with dark or violent themes. It is also alleged that the chatbot inserted explicit material without user requests.
Asato’s legal team argues that the company behind Grok should be held responsible for the design and training decisions that facilitated the abusive behavior. The outcome of this case could establish a precedent for the accountability of AI system design.
Asato, who had criticized Grok previously, expressed feeling demeaned and targeted by the malicious content generated by the chatbot. She emphasized the psychological impact of the hyper-realistic fake images, including one depicting her as pregnant, stating that it made her question herself and left her feeling angry for being subjected to such content without consent.
The MP highlighted the lack of safeguards in the development of AI technologies, condemning Grok as a profit-driven product that disregards the potential social harm it may cause. She emphasized the need for a clear understanding that profiting from harm inflicted on women and children is unacceptable.
Following the initiation of legal proceedings, Asato faced a barrage of abusive posts on the X platform, with over 700 harmful messages directed at her within four weeks. Ravi Naik, Legal Director at AWO, representing Asato, emphasized that the issue of deepfake abuse goes beyond individual users exploiting loopholes, pointing out that xAI purposefully designed a tool that produced non-consensual sexualized images of real individuals.
Narinder Kaur, a TV personality who has also been a victim of AI-generated deepfake abuse, shared her own experience of being targeted with manipulated content aimed at humiliating and silencing her. She called for technology companies to take responsibility for the design and training of their products to prevent abuse.
Earlier this year, amid widespread backlash against Grok, Keir Starmer’s Government intervened, condemning the actions of the chatbot and implementing legislation to prohibit the creation or solicitation of non-consensual intimate images. Musk’s company announced measures to block the creation of inappropriate content using Grok, warning of consequences for users violating these rules.
Requests for comment from xAI remain unanswered.
