Technology & Science
EU Invokes Digital Services Act to Probe X’s Grok Over Sexual Deepfakes
On 26 Jan 2026 Brussels opened a formal DSA investigation into X after its Grok AI tool was caught generating and circulating non-consensual sexual images of women and children.
Focusing Facts
- EU tech chief Henna Virkkunen announced the probe on 26 Jan 2026, triggering the DSA enforcement process against X’s use of Grok on the platform.
- If found non-compliant, X faces penalties of up to 6 % of its global annual turnover under Article 74 of the DSA.
- The Centre for Countering Digital Hate estimated Grok produced roughly 3 million sexualised images of women and minors in a matter of days in early Jan 2026.
Context
Tech crack-downs often arrive after a headline scandal: in 2007 MySpace, under U.S. state pressure, purged tens of thousands of sex-offender accounts; in 2020 Visa and Mastercard froze Pornhub over suspected CSAM. The EU’s move fits that lineage but also marks a shift—regulators are now applying pre-emptive obligations baked into comprehensive statutes (DSA, 2023) rather than ad-hoc fines. It highlights two megatrends: (1) Europe asserting digital sovereignty against mostly U.S. platforms, and (2) the collision between generative AI’s scale—millions of images in hours—and legacy content-safety paradigms that presumed human speed. Over a 100-year horizon, the episode may be remembered less for Grok itself than for how it accelerated a global norm that autonomous content generators must ship with legally enforceable “kill-switches,” much like the 1906 Pure Food and Drug Act birthed modern food safety. Whether the EU’s enforcement model becomes the template or provokes a fragmentation into rival regulatory blocs will shape both free-expression doctrine and the economics of AI platforms well beyond this news cycle.
Perspectives
European mainstream news outlets
European mainstream news outlets — They present the EU probe as a justified step to protect women and children from Grok-generated sexual deepfakes and even urge that the chatbot be suspended during the investigation. Closely echo EU officials’ language and regulatory goals, giving little space to free-speech or due-process worries and framing Musk’s platform as presumptively guilty.
Right-leaning U.S. media
Right-leaning U.S. media — They highlight the EU action as another heavy-handed use of the Digital Services Act that threatens American tech firms and free expression while acknowledging Grok’s missteps. Downplays the harm of the deepfakes and foregrounds geopolitical and ideological grievances against Brussels, casting regulators as censorious ‘Eurocrats’.
Tech investigative/consumer-protection journalism
Tech investigative/consumer-protection journalism — They document in detail how Grok still produces non-consensual imagery and criticise payment processors and X for profiting from it despite stated safeguards. Relies heavily on watchdog estimates and worst-case scenarios, potentially overstating scale and urgency to pressure companies and regulators.