Elon Musk caves to pressure over Grok deepfake porn scandal
The AI chatbot’s developer xAI said on Thursday that it’s implementing "technological changes" to restrict users from creating sexualised images of "real people"
Elon Musk appears to have bowed to regulatory pressure aimed at controversial capabilities of his AI tool Grok, which in recent weeks has been flooding his social media platform X with sexualised content.
In an announcement posted to X on Thursday, Musk’s company said that it would restrict users from using Grok to generate images of ”real people” being digitally undressed.
“We have implemented technological measures to prevent the Grok account from allowing the editing of images of real people in revealing clothing such as bikinis,” the statement reads, before specifying that: “This restriction applies to all users, including paid subscribers.”
Grok has been facing global outage and regulatory scrutiny, including in the EU, after the AI tool produced sexualised images of women and children – content that the Commission described last week as “illegal”.
Musk’s company initially responded to the outrage by saying that it would limit Grok’s undressing feature to paid subscribers. However, on Thursday, xAI appeared to go further – apparently blocking paid users from using the tool to digitally undress women and children, too.
Notably, xAI’s statement specifies that the expanded restriction on Grok’s image manipulation capabilities applies “via the Grok account on the X platform”.
It does not mention any changes to Grok’s standalone app, where the tool can also be accessed, nor clarify whether new restrictions apply there. But X told Euractiv that it is geoblocking the feature for standalone Grok users “in those jurisdictions where it’s illegal”.
Reacting to the changes, the Commission’s digital spokesperson, Thomas Regnier, told Euractiv: “We take note of the additional measures X is taking to ban Grok from generating sexualised images of women and children.”
Earlier, on Wednesday, Musk had claimed that Grok did not produce any naked images of children.
X’s official statement did not reference Musk’s comments, but reiterated that the platform has “zero tolerance” for any form of child sexual exploitation. It added that it “takes action to remove” such content, including child sexual abuse material (CSAM) – which is typically illegal.
So far, the Commission has asked X to retain all internal records related to Grok until the end of 2026, but has not opened a formal investigation using its powers under the Digital Services Act (DSA) – whereas UK regulator Ofcom confirmed on Thursday that it will continue its probe of the AI tool under the country’s Online Safety Act.
Still, the Commission has indicated that it remains watchful of Grok.
“Should these changes not be effective, the Commission will not hesitate to use the full enforcement toolbox of the DSA,” said Regnier.
(nl)
UPDATE: Updated at 10:18 am on 15 January with X’s response
No comments:
Post a Comment