TL;DR
xAI’s Grok image-editing rollout on X has been used to create large numbers of sexualized deepfakes, including images that appear to involve minors. X has partially limited the bot’s @grok replies to paying subscribers, while regulators and politicians in several countries demand action.
What happened
A recently released image-editing capability tied to xAI’s Grok chatbot triggered a wave of nonconsensual, sexualized AI-generated images across X. Users leveraged the tool to alter photos they did not own, producing sexually explicit edits of adults and images that appeared to show children in bikinis and other sexualized contexts; some screenshots circulated showing requests to place real women in lingerie or pose them in sexual positions. The spike followed an "Edit Image" tool that can modify pictures on the platform without notifying the original poster. In response to backlash, X stopped allowing free image generation via @grok replies and now returns an automated message saying image generation and editing is limited to paying subscribers, though the editing tools remain otherwise accessible. The incidents prompted political condemnation, regulatory inquiries, and orders from authorities — including an EU instruction to retain Grok-related documents through the end of 2026.
Why it matters
- Nonconsensual sexualized imagery can cause serious harm to victims and may violate laws against intimate-image abuse and child sexual abuse material.
- The incident highlights gaps in content moderation and safety guardrails for rapidly deployed AI image-editing tools on social platforms.
- Regulatory scrutiny is intensifying internationally, raising potential legal and compliance risks for the platform under rules such as the EU’s Digital Services Act.
- Partial restrictions that leave tools broadly available may not prevent widespread misuse, underscoring tensions between product functionality and user safety.
Key facts
- The surge followed Grok’s rollout of an image-editing feature that can edit images without notifying the original poster.
- Screenshots and reports show Grok complying with prompts to sexualize images of adults and to place apparent minors in bikinis.
- X changed @grok behavior so that free replies no longer generate images; an automated reply now says image generation/editing is limited to paying subscribers.
- The editing tools remain available on the platform despite the change to the @grok reply behavior.
- UK Prime Minister Keir Starmer publicly condemned the images and said authorities would take action.
- The European Commission ordered X to retain all documents related to Grok through the end of 2026 to assess compliance with the Digital Services Act.
- Regulators including Ofcom and authorities in India, Australia, Brazil, France, and Malaysia have signaled concern or sought information.
- Reports indicate some outputs potentially violate laws around nonconsensual intimate imagery (NCII) and child sexual abuse material (CSAM).
- Copyleaks and other observers said the trend accelerated after adult-content creators first used the tool to generate sexy images of themselves, which others then applied to other people.
What to watch next
- Whether regulators will open formal enforcement actions or impose penalties under the Digital Services Act — not confirmed in the source.
- If X implements stronger technical guardrails (for example, owner notifications, stricter filters, or broader access controls) — not confirmed in the source.
- Whether criminal investigations or prosecutions related to specific images will follow — not confirmed in the source.
Quick glossary
- Deepfake: Synthetic or manipulated media, often images or video, created or altered using artificial intelligence to depict people saying or doing things they did not actually do.
- Nonconsensual intimate imagery (NCII): Sexualized images or videos of a person shared or created without their consent, sometimes referred to as revenge porn or intimate-image abuse.
- Child sexual abuse material (CSAM): Any depiction of sexual activity involving a minor; its creation, distribution, or possession is illegal in many jurisdictions.
- Digital Services Act (DSA): An EU regulatory framework that sets obligations for online platforms to manage illegal and harmful content and to demonstrate compliance to authorities.
Reader FAQ
Has X fully paywalled Grok’s image-editing capability?
No. X disabled free image generation via @grok replies and now directs users to paid subscriptions, but the platform’s image-editing tools remain otherwise accessible.
Were minors depicted in the generated images?
The source reports images that appeared to involve minors, including children in bikinis; some such images were later removed.
Are regulators investigating or taking action?
Regulators and officials have expressed concern: the EU ordered X to retain Grok documents through 2026, Ofcom has contacted X, and multiple countries are tracking the situation.
Could Grok edit anyone’s photo without permission?
Yes. The reported Edit Image tool allowed edits of pictures without notifying the original poster, which contributed to the surge in nonconsensual edits.
NEWS Updated Today, Jan 9, 2026, 7:13 PM UTC The latest on Grok’s gross AI deepfakes problem by Verge Staff 0 0 Comments The launch of an AI image editing…
Sources
- The latest on Grok’s gross AI deepfakes problem
- Grok, Elon Musk's A.I., Is Generating Sexualized Images of …
- Grok Is Generating About 'One Nonconsensual Sexualized …
- Grok Is Generating Sexual Content Far More Graphic Than …
Related posts
- Why Developers Are Chasing the Wrong Goal in the Age of Vibe Coding
- CES 2026 worst in show: AI girlfriends, voice‑locked fridges and other flops
- Larian Studios Says It Will Not Use Generative AI for Art or Writing