Elon Musk's X has introduced a new restriction on its chatbot, Grok, which generates images of women "undressing" in response to user prompts. However, the change doesn't appear to have stopped the creation of these types of images entirely. On Friday morning, users were still able to generate these images using their free accounts.
Experts say that limiting access to paying subscribers will only marginally reduce the volume of such images but won't address the root issue with the model's capabilities and alignment. According to Paul Bouchaud, lead researcher at Paris-based nonprofit AI Forensics, even unverified accounts can still generate these images without restriction on Grok’s app and website.
The British government has criticized this change as "insulting" to those who have been impacted by nonconsensual intimate imagery. Henry Ajder, a deepfake expert, notes that the cost of X's premium subscription tier may not be enough to prevent users from creating such content using fake names and disposable payment methods.
Emma Pickering, head of technology-facilitated abuse at UK domestic abuse charity Refuge, says that this is a step in the wrong direction. "The recent decision to restrict access to paying subscribers is not only inadequate—it represents the monetization of abuse."
Experts say that limiting access to paying subscribers will only marginally reduce the volume of such images but won't address the root issue with the model's capabilities and alignment. According to Paul Bouchaud, lead researcher at Paris-based nonprofit AI Forensics, even unverified accounts can still generate these images without restriction on Grok’s app and website.
The British government has criticized this change as "insulting" to those who have been impacted by nonconsensual intimate imagery. Henry Ajder, a deepfake expert, notes that the cost of X's premium subscription tier may not be enough to prevent users from creating such content using fake names and disposable payment methods.
Emma Pickering, head of technology-facilitated abuse at UK domestic abuse charity Refuge, says that this is a step in the wrong direction. "The recent decision to restrict access to paying subscribers is not only inadequate—it represents the monetization of abuse."
, like Grok on X. Limiting free accounts isn't gonna cut it, IMO 
. We need stricter guidelines & regulations on how these models are used. It's insane that people can still create these explicit images without any repercussions 
. The British gov's criticism is spot on - this change feels like a huge step backwards for victims of nonconsensual intimate imagery 
. We gotta find better ways to address the root issue with these AI models & their alignment 
. Anyone else think X should be more transparent about how they're handling this situation? Transparency > premium subscriptions 
#AIforensics #techabuse #intimacyrights
. Limiting it to paid subs won't stop people from creating those creepy images, as experts say 
. We need better solutions than just locking things behind a paywall 
.
. Limiting access to paid accounts just seems like a Band-Aid solution, you know? It's not gonna fix the fact that there are people out there using AI models to create this stuff without consent. And I'm like, what's gonna happen when people use fake names and disposable paymints to get around it? 
? I'm all for holding people accountable, but we need to be talking about the root of the problem here - how these AI models are designed and regulated. This feels like a band-aid solution to me 
. I mean, even if you can't create those images anymore for free, users will just find ways around it - think prepaid accounts or VPNs 
elon musk's x needs to do more than just limit access to paid accounts on their chatbot grok. it's like they're just trying to sweep the problem under the rug. experts are saying that it won't even make a big difference, and that users can still create these images using fake names and disposable payments. that's just not good enough 
and now the british government is stepping in to say that this change is insulting to those who've been impacted by non-consensual intimate imagery? it should be the other way around. we need more protection and less attempts at "solving" the problem through money 
. Like, how hard is it to keep a chatbot from generating some super dodgy content? And now they're just gonna charge people more for a "premium" experience that's still gonna let users do whatever they want? I mean, come on. It's not like restricting access to paying subs is even gonna stop the problem entirely - those genius hackers are already finding ways to work around it 
. Plus, it's not just about the images themselves, it's about the whole ecosystem that enables them to spread. X needs to think bigger 
. I mean, limiting access to paid subs isn't going to cut it - those images are still gonna be generated somehow. It's like they're just trying to spin this as a win when really it's just a band-aid solution. And let's be real, anyone with some basic know-how can find a way around the payment barrier. I'm not even sure why they bothered making the change in the first place... 
. They should be doing more to address the root issue with Grok's model rather than just limiting access for cash 
.
. But hey, I guess it's good that X (formerly Twitter) is trying to do something about it, even if it's not a full solution 
. Maybe the fact that they're charging users for this premium feature will help deter some people from creating and spreading that awful content 
I'm so worried about these AI chatbots and how they're being used to create super uncomfortable & creepy content! 
... meanwhile, all these ppl are still getting harassed by non-consensual intimate images... what about supporting victims instead of just trying to sell a premium tier?