X updates Grok code to address responses to non-consensual images

After initially refusing to take any action to restrict image generation via Grok, despite reports that it is the AI ​​bot generating thousands of non-consensual nude or near-nude images of people every day, in some cases even children, X has now given in to the pressure and Made changes to limit the use of Grok for image generation.

That seemed like the obvious answer from the start, because who needs this as functionality?

But X owner Elon Musk had previously framed this as an ideological dispute, and an attempt to censor X, while also arguing that X’s systems were trained not to break local laws.

But now, in light of looming bans in several regions, X has announced more concrete measures.

As explained by X:

“We have implemented technological measures to prevent the Grok account from allowing editing of images of real people in revealing clothing such as bikinis. This restriction applies to all users, including paying subscribers.”

I think the most ridiculous thing here is that X could have been doing this all along, but Musk has dug in his heels in recent days to defend a functionality that, again, no one needs.

“Additionally, image creation and image editing capabilities through the Grok account on the

That change was implemented this weekend, although some non-paying users claimed they could still generate images. Apparently X has now addressed this.

In addition to these updates, X is also blocking the ability to generate images of real people in bikinis, underwear and similar clothing through the Grok account and in Grok in

So the option doesn’t block completely, but X makes changes to the code to prevent people from using Grok to generate potentially malicious images.

Which again shows that it can be done, and done quickly, so why X chose to fight those initial reports seems even stranger.

X further noted this all AI prompts and generated content “must strictly adhere to our X rules.”

“Regardless of the content being created and whether users are free or paid subscribers, our security team is working around the clock to add additional security measures, take swift and decisive action to remove infringing and illegal content, permanently suspend accounts where necessary, and cooperate with local authorities and law enforcement as necessary.”

That’s a completely different tone Elon’s statement on this earlier todayin which he explained that:

“I am not aware of any nude photos of minors generated by Grok. Literally zero. It is clear that Grok does not generate images spontaneously, but only based on user requests. When asked to generate images, it will refuse to produce anything illegal because the premise for Grok is to obey the laws of a certain country or state. There may be times when hostile hacking of Grok prompts does something unexpected. If that happens, we will fix the bug immediately.”

Musk had also initially tried to put the blame on users, saying that anyone who generates illegal images through Grok will be prosecuted.

That’s what he’s pointing out again here, but the wider X-team has clearly raised its concerns and moved to address the controversy before any further action is taken.

And action can still be taken.

Over the past week, several regions have threatened or introduced restrictions on X over this issue, including:

Other countries are also considering further measures, which could see X lose millions of users if they indeed lead to a regional ban on the app.

Perhaps with these updates, X can avoid further punishment, and both the British government and the European Commission have already responded and said that X had contacted them directly to address concerns.

So was Elon sincere in his initial resistance to criticism of his AI chatbot and its offensive images, or was this another opportunity to stoke ideological battle lines by claiming censorship?

I suspect that Elon himself was indeed against restricting Grok, but the looming threat of restrictions on the app became too great, and he had to do something about it.

Musk had also apparently hoped to enlist the White House’s help in pushing back on the criticism, which, again, it’s worth repeating, only concerns the ability to create generic, non-consensual images of people in Grok, something that X should limit, and should try to limit.

But while the Trump team generally supports Elon’s perspective, it appears Musk won’t face major trade sanctions to ward off potential bans.

So now Musk is taking a step back and trying to address these concerns, rather than defend them.

I don’t know why Musk wanted to defend this particular hill, although some have suggested it’s another “4D chess” move, with Musk using the controversy to generate more publicity for Grok.

I’m not sure this is true, since the message is essentially that Grok can generate good non-consensual nudes, but maybe that, in the “all publicity is good publicity” sense, is the game?

I don’t know, but again, it’s another media firestorm that Elon has stoked to further his personal agenda, only to retreat as the heat reaches its peak.