Mounting scrutiny into Grok’s ill-fated viral bikini image deepfake trend on X.com is really damning, when you look at the sheer scale of AI’s damaging effect. Bottomline is that Grok didn’t just “enable” a deepfake bikini trend on X – it accelerated it to scary levels.
Bloomberg’s latest reporting shows the AI bot was generating “undressed” image edits at a scale measured in thousands per hour on X.com, turning non-consensual, sexualised image manipulation into a public spectacle of epic proportions. And because it happens right inside the social media feed, the abuse was wrapped in bad faith humour and engagement like never before.
According to independent researchers who monitored Groq’s deepfake image outputs on X.com specifically between January 5th and 6th, at its peak the xAI powered chatbot was churning out manipulated, sexualised edits at a frenzied pace – about 6700 images generated per hour, reported by Bloomberg.
What added insult to injury was X.com’s role in socially amplifying the damage caused by Grok’s unchecked viral bikini trend. Because it didn’t happen on an obscure, dark corner of the web – it was a mainstream timeline feature on one of the world’s most visited online destinations.
Anyone who wanted to join in on the viral bikini trend simply had to reply to someone’s photo post on X, ask Grok to “edit” and let the engagement machine do the rest. Additionally, Reuters’ analysis also found cases where Grok generated sexualised images involving children – showing us all a global platform-scale safety failure.
In all of this, governance looked optional. X’s guardrails and enforcement appeared inconsistent, and the tone from the top didn’t help, as even Elon Musk responded with laughing emojis to bikini edits – including of himself – as the controversy spread.
India’s Ministry of Electronics and IT has issued a notice to X, flagging Grok’s misuse for obscene and sexually explicit content, demanding removals, and seeking an action-taken report under the IT Act and IT Rules, 2021.
Elsewhere, scrutiny is stacking up. Australia’s eSafety Commissioner is investigating Grok-generated sexualised deepfakes, including reports involving minors. In the UK, the Commons Women and Equalities Committee has said it will stop using X after the Grok image row, according to a report by The Guardian.
Grok didn’t invent non-consensual sexualised imagery, but its moment of AI reckoning is warranted. What it did along with X.com was lower the cost of harm to near zero, and raise the cost of being a target to something you can’t opt out of. If Grok (and X) want to claim they’re serious about “free speech,” as Elon Musk argues and fight for so passionately, they’re going to have to prove they’re equally serious about consent.
Also read: Grok vs Indian Govt: Why Musk’s AI is facing serious scrutiny in India
