Thread by @ChrisO_wiki

1/ Elon Musk faces a spiralling worldwide crisis and growing legal jeopardy over child pornography and nonconsensual sexual images being generated through his Grok AI chatbot. Multiple countries and jurisdictions have now announced investigations into X and xAI. ⬇️

2/ As reported by Reuters, many thousands of AI-generated sexual images have appeared on X over the Christmas and New Year period. They include images of real women being digitally undressed, repositioned in sexual poses, and covered in simulated semen ('donut glaze').

3/ The images created through Grok by X users have also reportedly included sexualised images of pre-teen children as young as four years old.

https://x.com/stclairashley/status/2008050483600973860?s=20

4/ According to the Paris-based AI Forensics group, at least a quarter of Grok's 50,000 mentions on X between 25 December and 1 January were requests to create an image, with a high prevalence of terms such as “her”, “put”, “remove”, “bikini” and “clothing”.

5/ More than half the 20,000 images generated over that period were of minimally dressed women, with a minority of the images, or 2%, appearing to show people aged 18 or under. Some represented children of five years old or younger.

6/ This has prompted widespread complaints about Grok, particularly from women whose pictures have been appropriated to sexualise them. More ominously for X, multiple jurisdictions have announced investigations and crackdowns. They include:

7/ 🇪🇺 European Union: The EU is "very seriously looking" into complaints about Grok's production of sexual images. Digital affairs spokesman Thomas Regnier says that "This is illegal. This is appalling."

8/ 🇫🇷 France: The Paris prosecutor's office has widened an existing investigation into X, noting that sexual image offences are punishable by two years' imprisonment and a €60,000 fine. X is already under investigation for disseminating antisemitic and Holocaust denial content.

9/ 🇮🇳 India: The Ministry of Electronics and Information Technology has ordered X to block the creation and dissemination of AI-generated material involving nudity, sexualisation, or other unlawful content, and to submit a compliance report within 72 hours.

10/ 🇲🇾 Malaysia: The Communications and Multimedia Commission says it has “taken note with serious concern of public complaints about the misuse of artificial intelligence (AI) tools on the X platform, specifically the digital manipulation of images of women and minors…

11/ …to produce indecent, grossly offensive, and otherwise harmful content.” It adds that it is “presently investigating the online harms in X.”

12/ 🇬🇧 United Kingdom: Communications regulator Ofcom has announced that it has made urgent contact with X and xAI to "undertake a swift assessment to determine whether there are potential compliance issues that warrant investigation."

https://x.com/Ofcom/status/2008201578378084550?s=20

13/ Further investigations around the world are almost certain to be announced in the coming days as the controversy continues and political pressure mounts for action.

14/ Grok itself has said that "There are isolated cases where users prompted for and received AI images depicting minors in minimal clothing... xAI has safeguards, but improvements are ongoing to block such requests entirely."

https://x.com/grok/status/2006618883055046758?s=20

15/ The root cause of the problem appears to be an update rolled out to Grok in December 2025 that offered a new 'spicy mode' by which sexualised content could be generated. So-called 'nudifiers' are not new, but the Grok update has made them much more more accessible.

16/ The commentary below by @ednewtonrex is well worth reading. As he points out, "There is ~zero chance this didn't come up in a meeting at X ahead of release. (Or, if it didn't, something has gone seriously wrong at the company.)"

https://x.com/ednewtonrex/status/2008098762820923688

17/ Probably not unrelatedly, Elon Musk initially reacted with laughing emojis at some of the images being generated. He subsequently posted the following after the global outcry began:

https://x.com/elonmusk/status/2007475612949102943

18/ An X spokesperson has also said: “We take action against illegal content on X, including child sexual abuse material (CSAM), by removing it, permanently suspending accounts, and working with local governments and law enforcement as necessary.”

19/ However, reporters say that abusive images have remained up even as the accounts which requested them have been banned.

20/ Grok's new mode poses serious legal liabilities for X and xAI. In the EU, both companies likely face liabilities under the Digital Services Act (DSA). Individual members states' laws against deepfakes and child sexual abuse material are also likely to be implicated.

21/ In the US, every state has laws against child sexual abuse material, which state attorneys general may use against X/xAI. Congress also passed the TAKE IT DOWN Act in 2025, which prohibits the nonconsensual publication of intimate visual depictions, including deepfakes.

22/ X/xAI may seek to use Section 230 of the US Communications Decency Act to shield it from liability for content posted by their users. However, Grok is not a user, but a service provided by the platform itself – so it may not have any cover under S230.

23/ It's likely that X/xAI will roll back Grok's 'spicy mode' in the face of growing legal and political pressure. However, the damage may already have been done in terms of legal exposure for the two companies and harm for Grok's many victims. /end

Sources:
🔹 reuters.com/legal/litigati…
🔹 theaiinsider.tech/2026/01/05/ind…
🔹 lemonde.fr/en/internation…
🔹 politico.eu/article/france…
🔹 theguardian.com/technology/202…
🔹 facebook.com/photo/?fbid=14…