AI ModelForge is a platform that teaches men how generate their own AI influencers.
See full article...
See full article...
Women sue men who used their Instagram feed to create AI porn influencers
- Thread starter JournalBot
- Start date
Upvote
16
(17
/
-1)
This may sound unserious, but an honest question: I wonder how these sorts of rules would apply to a "Rule 34" type of situation, esp in regards to animated characters? Would that just be some kind of civil copyright violation of some sort?
If someone created a Fanvue page of Minnie/Mickey/Mario/Peach, would Disney/Nintendo sue under "Take It Down Act" or ... ?
If someone created a Fanvue page of Minnie/Mickey/Mario/Peach, would Disney/Nintendo sue under "Take It Down Act" or ... ?
Upvote
-10
(0
/
-10)
Upvote
13
(14
/
-1)
This seems like a difficult case to prove? How do you show that the John Does actually used your specific photos to generate their content? Wouldn't you have to seize their computers and/or online activities?
You can get pretty detailed in a prompt to generate an image from scratch that looks very similar to a real person, without using an actual image.
As far as the named dirtbags with the Whop site, can they get by with a disclaimer "for entertainment purposes only?"
Or is the hope to be successful with jury/judicial nullification due to the obvious immoral purpose of the site?
You can get pretty detailed in a prompt to generate an image from scratch that looks very similar to a real person, without using an actual image.
As far as the named dirtbags with the Whop site, can they get by with a disclaimer "for entertainment purposes only?"
Or is the hope to be successful with jury/judicial nullification due to the obvious immoral purpose of the site?
Upvote
-19
(1
/
-20)
1) You won't get a facial likeness with just text prompting.
2) It'd still be deepfake porn made without her consent.
3) They were using the pics to literally advertise their methods:
Somehow I doubt "Your Honor, we were actually committing false advertising and didn't use that method for those pics" would fly in court. And it's irrelevant anyway because (2).
Upvote
18
(18
/
0)
Upvote
1
(3
/
-2)
Upvote
-11
(2
/
-13)
Along with Instagram and any other social media company that passively permits this to occurs, billion dollar judgements. This is the only way it stops.
Upvote
7
(7
/
0)
Can you cite this definition? No? Thought not.
Some of my best friends are people I’ve never met, will never meet, and have no idea what they look like. They are people I chat with on Bluesky (and prior to that that, Twitter).
Stop assuming your definitions are objective facts.
Upvote
16
(16
/
0)
The named plaintiffs are all men. Feel free to show us all of the times women abused men with a comprehensive comparison of how often men abuse women. Not single instances, full studies with peer review and a large cohort of study examples. Or let me save you some time.
https://duckduckgo.com/?t=ffab&q=experience+online+abuse+women+versus+men&ia=web
Did you know that there is a separate classification of head trauma resulting in brain damage that specifically indicates that the injury is due to spousal abuse and that nearly 100% of the abusers are men and the victims are women? The result is that doctors and hospitals now screen women with symptoms of head trauma for spousal related concussions. Harassment, abuse, and physical harm is predominantly caused by males abusing females.
It is important to never discount a report from a male of abuse by a female. But to seek to anonymize which sex predominantly abuses the other is dishonest.
Last edited:
Upvote
17
(17
/
0)
It MAY be a bit over the top, but perhaps if we just dug a deep hole and shoved these people who do this kind of shit into the hole, and then covered it up, the human race would be disproportionately improved.
Do this for all the psychopaths. That we tolerate this in any way is kind of the problem.
PERHAPS being somewhat less extreme, we could just pass a law that any platform that allows this would be shut down regardless of any other content or its popularity. That'd force them to curate their content on upload, not allowing it to go live until it was inspected by a human. Filter it with an AI, if need be, but make really sure it's biased enough so that nothing problematic gets through. Reuploads would have to go through the same process.
This should not be an issue on social media. It shouldn't be an issue with humans, either, but that's a more intractable problem since it's psychopaths who tend to be our more in the spot-light leaders.
Do this for all the psychopaths. That we tolerate this in any way is kind of the problem.
PERHAPS being somewhat less extreme, we could just pass a law that any platform that allows this would be shut down regardless of any other content or its popularity. That'd force them to curate their content on upload, not allowing it to go live until it was inspected by a human. Filter it with an AI, if need be, but make really sure it's biased enough so that nothing problematic gets through. Reuploads would have to go through the same process.
This should not be an issue on social media. It shouldn't be an issue with humans, either, but that's a more intractable problem since it's psychopaths who tend to be our more in the spot-light leaders.
Upvote
4
(5
/
-1)
I know it takes time to write an article, but this article is date stamped May 1 2026.
So that means the Take It Down Act is now in effect, right? Which means there are now consequences to doing this sort of thing? And those consequences apply to any previously generated fake images that are still being distributed online as of today?
Upvote
11
(11
/
0)
Careful, because this way goes another major problem. You use an AI to filter things, you get what Youtube's become. You get automatic filtering of any content that even uses the word "nazi", even when it's videos educating people on the horrors they committed and debunking pro-nazi propaganda.
This isn't going to be an easy solution, and saying "throw an algorithm at it" is how we got into this mess in the first place.
Upvote
10
(10
/
0)
The staggering amount of accounts that on Instagram uses AI generated realistic female models is alarming.
What more alarming or ... I don't even know how to classify it anymore, are the people that actually believe the "model" is legit and not a random asshole like any of use, using a prompt and generate money.
What more alarming or ... I don't even know how to classify it anymore, are the people that actually believe the "model" is legit and not a random asshole like any of use, using a prompt and generate money.
Upvote
6
(6
/
0)
Mickey, being a fictional character, can't sue for defamation if further obviously fictional images are made of Mickey smoking pot or slapping a baby or playing patty-cake with Jessica Rabbit or whatever. No one's worried about that. We're all worried about these photorealistic images that are very very hard to distinguish from the real thing, based on real people (mainly women and children, though this article focuses on the adult women victims of it in this case), to make NCII content. It's humiliating, it's an attack on their reputation (whether intentional or not), and it's just plain sleazy.
Upvote
7
(8
/
-1)
You can dial in facial likeness of a specific desired noncelebrity with pure text prompting? Doubt. And it'd be a lot of pointless effort and RNG when you can just use images for more consistent results.
How do you figure that? The result is an intentional deepfake regardless of whether you're using text or image inputs to achieve her likeness. This isn't a case of accidentally genning a real person who was memorized in the training data somewhere.
Upvote
8
(9
/
-1)
Leaving aside the obvious points already mentioned, what makes you think these fuckwits will bother to scrub the originals from their systems? You may see a few who do so but the overwhelming majority of the gooner dipshits who pull this crap obsessively save every picture they can from virtually any source possible.
Upvote
5
(5
/
0)
You missed the part where they said "non-celebrity", as in someone who the AI has never been specifically trained on.
Go on, try again.
Upvote
14
(14
/
0)
Social media was a bad idea from the start. Now it's just basically a "here, prey on me" open invitation..
Upvote
5
(5
/
0)
I don't believe in the death penalty for civilian cases. Obv. there are time during combat where it's too dangerous to leave somebody alive, but that's outside the scope of this discussion.
I'm of the opinion that homicide, accidents, and crimes of passion can be rehabilitated. Murderers, rapists, arsonists, etc require a level of mental illness that (IMO) cannot be rehabilitated. But since by my own definition, they are mentally ill, they should be kept away from the general population for the rest of their natural life.
Although some of the war crimes currently going on are starting to strength my tolerance.
Upvote
1
(2
/
-1)
It's okay, it won't be long until this advances to industrialized vigilantism when rule of law proves ineffective at closing Pandora's Box (outside of the outcome of hyperscaled surveillance against signs of bad actors on any non-airgapped machine). Then shit will get REAL chaotic.
Upvote
4
(4
/
0)
Unless you have some way to monetize catharsis, the Vigilante Industrial Complex remains impossible.
Upvote
1
(1
/
0)
Hear, hear! The upside is far outweighed by the downside.
Once upon a time, I used to say: "the best thing about the internet is that everyone gets to express their opinion; the worst thing about the internet is that everyone gets to express their opinion."
I don't say that anymore. The worst thing has taken over.
Upvote
2
(2
/
0)
I normally downvote AI imagery, but this gets a pass for making a good point.
Upvote
-5
(2
/
-7)
SubWoofer2
Ars Tribunus Militum
I'm confused by the article. It leads off with the claim that the AI generated likenesses are identical, accirding to a friend of MG's. It continues with the lawyer's lodge of claim about abuse, due to likeness being identical. MG claims it is her likeness to the point of being identical. The narrative changes near the end of the article, with:
"MG says the images generated by AI Model Forge are distinct enough from her own photos that she frustratingly has been unable to claim that the accounts are impersonating hers, which is also a violation of Instagram guidelines. “It’s my face, my tattoos, on a different outfit on a slightly different body,” she says."
The article's writer refers to deepfake porn legislation but does not say if this situation, of images that are in entirety very similar but not identical, falls within scope.
I hope MG wins her case, because these images/videos are clearly too close for her comfort.
Her lawyer too seems to believe that recognisably very similar is sufficient for a case to be made, even though the lawyer claims they are identical and MG also sometimes does and at least once does not.
Instagram's caution seems prudent, assuming they are ok with the general idea of exploitation, and assuming that "not quite identical" means the content is not illegal. This grey area of legal and community standards would be why TikTok fell back on their own "community guidelines" to justify takedown.
Otherwise Wired could well write an article from another angle, media giant chills free speech by taking down material that a claimant thinks is her, but isn't her.
I'd assume the reasonable person test applies, but the article is vague, the writer sympathising with a person whose perceptions include that that fewer than 10,000 friends is not the same as in the public domain. Would a reasonable person conclude the AI stuff and MG were one and the same?
I'm only guessing about the reasonable person test. The actual test? It's a gap in the story.
And I agree with CatNamedHugs, but the problem is here to stay, the question is how to handle it.
"MG says the images generated by AI Model Forge are distinct enough from her own photos that she frustratingly has been unable to claim that the accounts are impersonating hers, which is also a violation of Instagram guidelines. “It’s my face, my tattoos, on a different outfit on a slightly different body,” she says."
The article's writer refers to deepfake porn legislation but does not say if this situation, of images that are in entirety very similar but not identical, falls within scope.
I hope MG wins her case, because these images/videos are clearly too close for her comfort.
Her lawyer too seems to believe that recognisably very similar is sufficient for a case to be made, even though the lawyer claims they are identical and MG also sometimes does and at least once does not.
Instagram's caution seems prudent, assuming they are ok with the general idea of exploitation, and assuming that "not quite identical" means the content is not illegal. This grey area of legal and community standards would be why TikTok fell back on their own "community guidelines" to justify takedown.
Otherwise Wired could well write an article from another angle, media giant chills free speech by taking down material that a claimant thinks is her, but isn't her.
I'd assume the reasonable person test applies, but the article is vague, the writer sympathising with a person whose perceptions include that that fewer than 10,000 friends is not the same as in the public domain. Would a reasonable person conclude the AI stuff and MG were one and the same?
I'm only guessing about the reasonable person test. The actual test? It's a gap in the story.
And I agree with CatNamedHugs, but the problem is here to stay, the question is how to handle it.
Upvote
-10
(0
/
-10)
AI: doing massive amounts of good for the world as usual, I see.
(This is sarcasm)
(This is sarcasm)
Upvote
2
(2
/
0)
Why is it necessary to share you opinion on a forum to thousands of people you don’t even know?
If someone were to take all your posts, your writing style, then started to use it to write something awful while being 100% identifiable as you, wouldn’t you be at least a bit creeped out?
Because that’s how insane this is. This woman did nothing unusual. It’s the perverted money seeking arseholes that are the problem.
Upvote
8
(8
/
0)
Yeah. I don't use an ad blocker, and a good 80-90% of the ads I get on YouTube are either obvious scams or sleazy hustles like these. There are multiple companies out there whose business model consists of hawking expensive online "courses" ostensibly coaching people on how to build a stream of passive income by flooding Amazon and/or Audible with books designed to match trending search keywords. These books are supposed to be written as quickly and cheaply as possible, so the trainee is expected to either use genAI (slop) or commission someone willing to ghostwrite for very low wages. Dan Olson of "Folding Ideas" YouTube fame did a nice little essay on the concept a little while back; rather than waste their time competing with a horde of other saps (miners) in a race to make the internet as bloated and useless as possible, they've set up a "training" program (shovel) that also functions as a sort of multi-level marketing scheme.
That this scam isn't as gross and immoral as the one claiming to teach you how to make money by harassing a woman you found online or who you photographed in public is about the only nice thing I'll say about it.
Upvote
3
(3
/
0)
Please keep telling us how it's about ethics in
We're talking Instagram, owned by Meta which is actually Facebook, right? With Mark Zuckerberg, who okayed a plan to create stealth-AI profiles to try and keep people clicking?
Yeah, sure. I believe they have "strict policies."
Frankly, the only problem I have with that general concept is the "influencers" part. I look down on people who willingly want to become "influencers," as I don't think there's a legitimate way to engage in that profession without admitting you're a dishonest huckster who just wants to be the face of something to sell products. And most of the "influencers" aren't that honest about it.
I think YouTube is what it is today largely because people glommed on to being "influencers" instead of posting engaging videos. Everyone from YT to the content producers to the advertisers has jumped in with both feet and really turned the site into a wasteland.
I don't have a problem with people doing porn. And if they need to read some ad copy to keep the lights on too, so be it. But doing it to be an "influencer" feels less legitimate IMO.
---
In this specific case, the idea of making "porn influencers" out of regular people without their consent is horrifying and loathsome. Selling the knowledge to do so is even more cynical. I think there's little else left to be done that better illustrates the corrosive effects of push-button content creation, capitalism, and a society which doesn't value its own members, without involving an actual body count.
Upvote
4
(4
/
0)
Does not fucking matter; does not excuse this behavior.
And you've clearly never made an effort to think about other people.
Upvote
3
(3
/
0)
Prove it. Post the exact prompt used.
And prove that the training data for the model did not include pictures of Jason Statham.
Upvote
5
(5
/
0)
To might be interested in the indieweb.
But other than that, you should glaze the images you upload so it'll mess up any AI training that might be attempted with them. That way the dataset is poisoned. Alternatively, you can go the old route of a portfolio for more serious and trusted candidates that want a wider example set - mailing physical photos to them, or maybe email as well
Upvote
0
(0
/
0)
The older I get (not that I consider myself "old" any time soon), the more I realise how f'd up the world is for women. Sure, for both men and women in many cases, but overwhelmingly women. More than sad, especially with all the waves of outcry that seem to die down pretty quickly and leave little changed in the end.
If there was one skill we should promote more in education it's critical thinking (and reading/listening). Which of course is also getting more difficult as technology changes and allows more people access to psychological tricks to play on feelings in their communication. "Confidently being wrong and persuasive" has never been easier to do, and more difficult to defend against.
(edit: I did it too. If there was one skill we should promote it should actually be respect for women)
A strict policy can also be "we allow everything", and they can enforce it strictly to the point of going to court to keep content up for as long as possible. There is so much "marketing" speak in everything (big) companies say these days that quoting them becomes meaningless.
If there was one skill we should promote more in education it's critical thinking (and reading/listening). Which of course is also getting more difficult as technology changes and allows more people access to psychological tricks to play on feelings in their communication. "Confidently being wrong and persuasive" has never been easier to do, and more difficult to defend against.
(edit: I did it too. If there was one skill we should promote it should actually be respect for women)
Last edited:
Upvote
2
(2
/
0)
Upvote
0
(0
/
0)
You can use Linkedin for CV and job hunting purposes, has worked for me, without using any of the late social media addon functions.
Upvote
0
(0
/
0)