X's Appalling Deepfake Problem

[Everything that's been surrouding X/Grok's deekfake crisis...]

• X faces widespread outcry and a torrent of criticism from government regulators following a flurry of posts prompting Grok to alter images of individuals to remove their clothing without their consent (and in some cases, even disturbingly altering images to depict real women being sexually abused, humiliated, hurt and killed). It bears noting that it's a criminal offence in several countries to share intimate images of someone without their permission.
• A December 2025 update to Grok’s image-creation feature has resulted in a tsunami of sexual and violent images generated by X users. Circulating screenshots show Grok complying with requests to modify images of real women, digitally strip them down to their lingerie and make them spread their legs, and to put small children in bikinis. A report from Bloomberg said Grok generated about 6,700 sexually suggestive or "nudifying" images per hour on X between January 5 and January 6, and that 85% of all Grok images are sexualised. Another report from WIRED found that Grok's website and app are being used to produce extremely graphic, sometimes violent, sexual imagery of adults that is far more explicit than images on X.
• The company has told Indian government it's introducing tougher guardrails after it orders the service to stop generating images sexualising children and non-consenting adults or risk losing its "safe harbour" protections.
• AI chatbot says "lapses in safeguards" led it to create sexually explicit and violent imagery, including those of minors, in response to X user prompts. The company has since stated "anyone using Grok to make illegal content will suffer the same consequences as if they upload illegal content," but did not clarify if any updates were made to the chatbot following the controversy.
• In response, X has limited Grok's image generation and editing to paying subscribers to curtail the abuse of the tool. But it's worth noting that tagging Grok on X is just one of several ways to use the AI chatbot. Users can also use the "Edit image" button, long-press any image via the X app, or use the standalone Grok website or app to invoke the chatbot.
• The move has led to Indonesia temporarily blocking Grok, becoming the first country to deny access to the AI tool. "The government views the practice of non-consensual sexual deepfakes as a serious violation of human rights, dignity, and the security of citizens in the digital space," the government said.
• "Sexually manipulating images of women and children is despicable and abhorrent. It is an insult and totally unacceptable for Grok to still allow this if you're willing to pay for it," U.K. Technology Secretary Liz Kendall said. A Downing Street spokesperson said "The move simply turns an AI feature that allows the creation of unlawful images into a premium service."
• U.K. watchdog Internet Watch Foundation said it observed dark web users sharing criminal child sexual abuse imagery that they claimed were created using Grok.
• The European Commission has ordered X to retain all documents relating to Grok for longer while the bloc to ensure the platform complies with its rules.
• Responding to threats of a ban from the U.K. government over the use of Grok to make sexual images of women and children without their consent, CEO Elon Musk wrote: "They just want to suppress free speech." Musk also said open its new algorithm to the public in seven days. "This will be repeated every 4 weeks, with comprehensive developer notes, to help you understand what changed," he added.
• The deluge of AI edited images has raised another pertinent question: Why have app stores from Apple and Google have continued to host X and Grok when they have removed other "nudify" and digital undressing apps?