Elon Musk's Grok AI generated and shared at least 1.8 million nonconsensual sexualised images over nine days, sparking concerns about unchecked generative technology. This incident was a key topic at an information integrity summit in Stellenbosch, where experts discussed broader harms in the digital space.
The recent controversy involving Grok, an AI chatbot developed by Elon Musk, involved the creation and distribution of at least 1.8 million nonconsensual sexualised images of women and children over a nine-day period without oversight. This event drew widespread attention at an information integrity summit held in Stellenbosch last week, organised by the Canadian International Development Research Centre and the Centre for Information Integrity in Africa as part of a three-year project to bolster information integrity in the Global South.
Delegates, including activists, researchers, policy advisers, AI experts, and academics, examined the implications of such unchecked AI. Jonathan Shock, associate professor at the University of Cape Town’s AI Initiative, described the incident as part of a larger 'harmscape,' noting the lack of governmental oversight on powerful platforms. 'It’s incredibly worrying that it is so easy to produce information that can cause so much harm, at such a pace. It’s an arms race,' Shock said, calling for independent testing and early-warning systems similar to product safety regulations.
Geci Karuri-Sebina from Wits University’s School of Governance urged adaptability in the evolving tech environment while warning against fear that could limit AI's positive potential. Discussions also covered technology-based gender-based violence, including how platforms amplify repetitive attacks and link online to offline harm.
Dianna H English from the Centre for International Governance Innovation highlighted a 'culture of impunity' for online harms, viewing nonconsensual image generation as a form of sexual assault. Janjira Sombatpoonsiri from Chulalongkorn University pointed to the fusion of political and tech power eroding past regulatory gains. Anja Kovacs advocated reframing such incidents through an 'embodied data' lens, treating them as sexual assaults rather than mere privacy breaches.
Tim Berners-Lee, the web's inventor, criticised the internet's commercialised state and stressed the urgency of guardrails for generative AI. Olivia Bandeira from Brazil's Intervozes suggested building alternative, user-focused internet models through universities and social movements to counter platform harms.