📱“Grok, put her in a bikini.” Over the past few days, social media, particularly X, has been flooded with a slew of NSFW imagery of individuals — men, women, and even minors — semi-naked and dressed in a smorgasbord of revealing undergarments. They’re all deepfakes — and they’ve never been easier to generate. The culprit? Grok. The prompt? A five-word sentence. The aftermath? Catastrophic.

…And it’s happening right before our eyes. X users with public profiles have found themselves victims of what the public AI model with over 30 mn active monthly users is calling a “lapse in safeguards,” allowing other users to generate realistic non-consensual sexual imagery using published media by mentioning and prompting Grok in the replies — results shared publicly and instantaneously. In a matter of hours, feeds were spammed, Reuters reports.

“There are isolated cases where users prompted for and received AI images depicting minors in minimal clothing […] Improvements are ongoing to block such requests entirely,” Grok posted on X — but the deepfakes kept coming, and it didn’t take long for the alarms to sound. On Friday, French authorities began investigating these deepfakes after two lawmakers reported the matter to the prosecutor’s office, which noted that the offense is punishable by two years’ imprisonment and a EUR 60k fine, Politico reports. India followed suit, its IT Ministry issuing an order directing Musk to to take corrective action, according to TechCrunch.

For Musk, it was business as usual, with the b’naire reacting to his favorite deepfakes — including ones featuring Musk in a bikini himself — with laugh-cry emojis. “The platform has been allowing the creation and distribution of these images for months without taking any action and we have yet to see any challenge by regulators,” Durham University Law Professor Clare McGlynn told the BBC. “[X or Grok] could prevent these forms of abuse if they wanted to,” McGlynn added. It’s worth mentioning that XAI’s policy prohibits depicting likenesses of persons in explicit manners. Musk’s solution ? Essentially a verbal slap on the wrist with no real action noted.

Yet, would they want to put it to an end? In the days following the “safeguard lapses,” Grok and X topped app store charts and Google searches, Musk claims — and one can easily surmise why: sexsells. Could Grok’s latest security hiccup be a carefully calculated update to test the waters of AI-generated R-rated content? After all, OpenAI’s ChatGPT isn’t that far behind on the NSFW front, with the AI giant set to introduce adult content for “verified” adult users, the Guardian reported earlier in October.

What was once a frightening “what if” is now our dystopian reality. Deepfakes have become all too common — explicit or otherwise. Over 50% of online articles are AI generated, as are 20% of videos shown to YouTube users. “Deepfakes will become mainstream, and the threat will shift from reputational damage to direct monetization by bad actors,” market research company Forrester says in its 2026 Trust and Privacy Prediction Report.

Tags: