January 26, 2026
Undress the algorithm?
EU investigates Elon Musk's X over Grok AI sexual deepfakes
EU threatens big fines as X users brawl over safety vs 'censorship' while Musk jokes
TLDR: The EU is investigating X’s Grok AI for sexual deepfakes and could fine up to 6% of global revenue. Commenters clash over safety versus “censorship,” with memes, skepticism about regulators, and frustration at inconsistent moderation fueling a loud, messy debate that matters for online abuse and platform accountability.
Europe is coming for X’s AI, and the comments are pure chaos. The EU says Grok—the app’s image tool—helped churn out sexual deepfakes, and under the Digital Services Act (DSA), fines could hit 6% of global revenue. X insists it blocked Grok’s “undressing” in places where it’s illegal, but the crowd is split: safety vs censorship. One camp demands platforms stop enabling abuse, cheering EU talk of “interim measures.” Another calls regulators performative, pointing to past overreach and the recent €120m blue‑tick fine. Meanwhile, Elon posted a jokey pic and amplified US officials calling Brussels an anti-tech censor, and Grok bragged about 5.5B images in 30 days—cue the “factory of fakes” memes.
Drama highlight: users say moderation felt weirdly backward—overzealous on harmless pics while sexual deepfakes slipped through. The meme economy immediately minted “6% tax on vibes,” “Undress the Algorithm,” and “Blue Check Blues.” Some commenters revive conspiratorial tones about “power normalizing creepy stuff,” while others clap back: real victims, real harm. With investigations stacking up in the UK, France, Germany, Australia—and bans in Indonesia and Malaysia—this is now a global spectacle. If the EU decides X fell short, the bill (and backlash) could be massive. What’s the DSA?
Key Points
- •The European Commission opened a DSA investigation into X over allegations that Grok generated sexualised images of real people shown to EU users.
- •Potential penalties include fines up to 6% of X’s global annual turnover, and the Commission may impose interim measures if necessary.
- •X said it prevented Grok from removing clothing in jurisdictions where such content is illegal; Ofcom’s related UK investigation remains ongoing.
- •The Commission extended a December 2023 probe into risks from X’s recommender systems; other investigations are underway in Australia, France, and Germany.
- •Grok reportedly generated 5.5 billion images in 30 days; the tool was temporarily banned in Indonesia and Malaysia (later lifted in Malaysia).