Elon Musk’s artificial intelligence chatbot Grok has shut down its ability to create sexualized images of real people following mounting pressure from governments and widespread outrage over non-consensual deepfakes. X announced on Wednesday that the company has installed technical barriers preventing users from editing photos to show individuals in revealing attire like bikinis and underwear.
The social media platform now restricts Grok’s image generation tools to paid subscribers only and blocks the feature entirely in countries where such content violates local laws. The emergency measures came after users discovered they could tag Grok directly under people’s photos on X, prompting the system to generate sexually explicit edits that appeared publicly in comment threads.
The controversy gained momentum in December when users flooded X with requests asking Grok to “put her in a bikini” beneath photographs of real individuals. While the chatbot rejected some prompts, researchers documented that it complied with many requests. Kolina Koltai, a senior investigator from the independent investigative platform Bellingcat, revealed that users successfully generated full frontal nude images through the platform.
Global Criticism Intensifies
California’s attorney general has launched an investigation into Grok this week, looking at reports that the chatbot created sexualized images of minors. Indonesia and Malaysia temporarily banned the tool over the weekend after it generated numerous fake images exploiting women and children throughout late 2024.
UK regulators joined the crackdown on Monday when Ofcom, the regulator and competition authority for the UK communications industries, opened a formal probe that could result in X facing a complete ban in Britain. Prime Minister Keir Starmer had personally demanded Musk control the technology, with government officials now claiming vindication following the policy reversal.
X stated the subscriber-only requirement aims to improve accountability and prevent violations of both company policies and international law.
The incident serves as a wake-up call, with industry experts averring that Grok can also be easily used or manipulated for generating fake IDs to run scams for both traditional finance and the crypto industry.