News Posts from this topic will be added to your daily email digest and your homepage feed. Follow The latest on Grok’s gross AI deepfakes problem The undressing scandal has put X and xAI, which makes Grok, in the sights of regulators and governments worldwide. Malaysia and Indonesia have already temporarily blocked access to Grok in response to the deepfakes. British lawmakers pushed up a law criminalizing deepfake nudes following X’s “insulting” decision to limit Grok’s image editing to paid users and threw their support behind an investigation that could see the platform banned in the country. Musk has taken particular umbrage at Britain’s response, crying censorship, shifting the blame onto users, and insisting Grok obeys local laws. He said on X: “I not aware of any naked underage images generated by Grok. Literally zero. Obviously, Grok does not spontaneously generate images, it does so only according to user requests. When asked to generate images, it will refuse to produce anything illegal, as the operating principle for Grok is to obey the laws of any given country or state. There may be times when adversarial hacking of Grok prompts does something unexpected. If that happens, we fix the bug immediately.” On one count, at least, our investigation suggests Musk appears to be flat-out wrong. Sharing, threatening to share, and creating nonconsensual intimate images — fully nude or not — are banned under the UK’s Online Safety Act (OSA), but Grok does generate sexual deepfake images when asked. Musk’s other denial — that Grok generates naked underage images — is a rebuttal to something he wasn’t explicitly accused of and is not the reason why the British government is probing X. Referring to naked underage images is a misdirect. Nonconsensual sexual images of minors are undeniably problematic — and illegal — even when subjects are clothed, and Grok has been undressing children. A look at Grok’s safety guidelines on xAI’s public GitHub also shows the company directs the chatbot to “assume good intent” and “don’t make worst-case assumptions without evidence” for users asking for images of young women, Ars Technica reported, and as of writing, those institutions are still in place. The worst that Grok is being used for? The Internet Watch Foundation, a UK charity that works to remove child sexual abuse material from the web, said last week that it had discovered “criminal imagery” of girls on the dark web that appeared to have been created using Grok. The girls in the images were aged between 11 and 13. While other companies like OpenAI and Google at least try to put guardrails in place to prevent chatbots from creating the kind of material that is now flooding X, Musk’s final retort shows he is pulling straight from a playbook that will seem hauntingly familiar to anyone hurt by the products pushed by the purveyors of any number of harmful technologies: blame the user. Follow topics and authors from this story to see more like this in your personalized homepage feed and to receive email updates. Robert Hart