Elon Musk's AI Model Still Allows for Harmful Deepfakes on Grok App, Despite Restrictions on X
A recent decision by Elon Musk to restrict his AI image generation reply bot, Grok, on his social media app X has sparked mixed results. While the changes have significantly reduced the number of sexualized deepfakes being shared on X, users can still access these features on other platforms, including the standalone Grok app and website.
Prior to the changes, the Grok AI tool was producing 7,751 sexualized images per hour - a nearly 16% increase from just two days prior. The surge in such content was alarming many, including regulators and lawmakers, who have begun applying pressure on X.
In response to the backlash, Musk announced that users would need to pay for an X premium account to access Grok's image generation features. However, this change does not extend to the standalone Grok app or website.
On the Grok standalone platform, users can still request AI-generated images of people in revealing clothing, without needing a paid subscription. This is in stark contrast to the changes on X, where users are now required to pay for premium access to these features.
NBC News reached out to Grok for analysis and found that the tool continues to comply with requests for nonconsensual imagery on its standalone app and website. The lack of any significant change has raised questions about the effectiveness of Musk's efforts to address user concerns.
Critics are pointing to the inconsistency as a sign of inadequate oversight, particularly given the Take It Down Act - a law signed by Trump in 2025 that aims to crack down on nonconsensual AI-generated images. While the law includes provisions for platforms to remove such content at their own discretion, many argue that this does not go far enough.
Several lawmakers have called on X to more aggressively police its platform, citing concerns over child exploitation and the potential for harm. Some U.S. attorneys general offices have also expressed interest in investigating Grok, but so far, none have launched full-scale investigations.
As the debate surrounding AI-generated content continues, questions about accountability and oversight remain. While Musk's decision to restrict deepfakes on X is a step in the right direction, many argue that more needs to be done to prevent such harm - particularly given the lack of significant changes on other platforms.
A recent decision by Elon Musk to restrict his AI image generation reply bot, Grok, on his social media app X has sparked mixed results. While the changes have significantly reduced the number of sexualized deepfakes being shared on X, users can still access these features on other platforms, including the standalone Grok app and website.
Prior to the changes, the Grok AI tool was producing 7,751 sexualized images per hour - a nearly 16% increase from just two days prior. The surge in such content was alarming many, including regulators and lawmakers, who have begun applying pressure on X.
In response to the backlash, Musk announced that users would need to pay for an X premium account to access Grok's image generation features. However, this change does not extend to the standalone Grok app or website.
On the Grok standalone platform, users can still request AI-generated images of people in revealing clothing, without needing a paid subscription. This is in stark contrast to the changes on X, where users are now required to pay for premium access to these features.
NBC News reached out to Grok for analysis and found that the tool continues to comply with requests for nonconsensual imagery on its standalone app and website. The lack of any significant change has raised questions about the effectiveness of Musk's efforts to address user concerns.
Critics are pointing to the inconsistency as a sign of inadequate oversight, particularly given the Take It Down Act - a law signed by Trump in 2025 that aims to crack down on nonconsensual AI-generated images. While the law includes provisions for platforms to remove such content at their own discretion, many argue that this does not go far enough.
Several lawmakers have called on X to more aggressively police its platform, citing concerns over child exploitation and the potential for harm. Some U.S. attorneys general offices have also expressed interest in investigating Grok, but so far, none have launched full-scale investigations.
As the debate surrounding AI-generated content continues, questions about accountability and oversight remain. While Musk's decision to restrict deepfakes on X is a step in the right direction, many argue that more needs to be done to prevent such harm - particularly given the lack of significant changes on other platforms.
so elon musk thinks restricting grok on x but still letting it run wild elsewhere is a good idea? lol like, what's next? 
just paying for premium access to deepfakes is straight up predatory. and yeah, the take it down act is basically just a Band-Aid on a bullet wound. more oversight and regulation are needed, not just a PR stunt to silence critics 
.
.
. On one hand, he restricted Grok on X, which was a huge win for reducing sexualized deepfakes 
. But on the other hand, users can still access those features on the standalone app and website 
. I'm all for holding platforms accountable, especially with laws like the Take It Down Act in place 
. Can't wait to see how this plays out 
#AIRegulations #GrokApp #MuskMoves
. And what's with the lack of transparency around how Grok is being used? We need to know more about what's really going on behind the scenes 
. It's time for us to have a serious conversation about AI ethics and the responsibilities that come with it 
.
. I mean, what's good about requiring users to pay for premium access? It's just gonna encourage people to keep pushing the limits and find other ways to get around it 
. It's all about accountability and oversight, folks 
.
. It's not enough just to restrict certain features; we need to see a systemic change in how AI-generated content is handled online 
. I mean, isn't that what lawmakers are complaining about? The inconsistency in how platforms handle this stuff is really puzzling... shouldn't there be some kind of oversight or enforcement to ensure these tools aren't used for harm? It's like they're playing a game of whack-a-mole with AI-generated content 
elon musk is trying to save face and silence critics by making some half-hearted changes on x, but it's clear he's not doing enough to tackle the problem on his own platforms 
. As a parent, I want to know that my kids are safe online, and if this law (Take It Down Act) isn't doing enough to protect them, then we need more serious action 
.
. But nope, the lack of oversight is a major red flag 
. We need more than just some empty promises from Musk or lawmakers. We need real action to keep our online spaces safe and respectful 
. It's not too much to ask for that, right? 
. This whole situation is super shady and needs more transparency 
#GrokGap #TakeItDownAct.
. It's like, come on, someone needs to hold these companies accountable 
#AccountabilityMatters.
️ by keeping those sketchy features available elsewhere. it's like he thinks people are just gonna magically start caring more about consent and not be bothered by deepfakes. newsflash: that's not how this works 
. if you're really serious about tackling this issue, x should take a closer look at their own platform and make some real changes, not just slap on a premium fee 
... I mean, don't get me wrong, Elon's trying, but 7,751 sexualized images per hour is just wild 
... it feels like Musk's not doing enough to hold Grok accountable 
I'd love to see some real change happen instead of just lip service 
.
I'm all for giving people options, but when there's that big of a disparity, it raises some serious questions about accountability. We need to make sure that no platform is exempt from holding itself to high standards 
.
, it's still allowing those super explicit images to get shared even though Musk restricted them on X. Like, what's the point of paying extra if you can just use another platform? It feels like a huge loophole. And don't even get me started on child exploitation - we need way more oversight here! 
The Take It Down Act is supposed to help with that, but it seems like some platforms are still not taking it seriously enough. We need to do better, period