Rendered at 19:43:46 GMT+0000 (Coordinated Universal Time) with Cloudflare Workers.
matt-attack 3 hours ago [-]
And shouldn’t we be encouraging AI generated CSAM? Who exactly is the victim when someone draws a picture, paints a picture or uses AI to do the same?
What kind of dystopia do we live in where there are illegal DRAWINGS?
Hugsbox 2 hours ago [-]
I don't know about "encouraging", but it is definitely a more complex issue than some people give it credit for.
intended 14 hours ago [-]
Yep. Several orgs have classifiers which are available now, NCMEC, Thorn, I believe the UK had one approved last year as well.
At a conference I was reminded out how prevalent it is in India, and how disturbingly cheap the cost of images is. That the team who handles it from the government is disturbingly small.
There are YouTube ads I saw in the past week, for tiny cameras. This was sold as a camera solution, except the video didn’t show what you would expect from home security or something like GoPro.
The borderline stuff had beautiful women putting cameras in some unusual locations, Like near the break pedals in a car, pointing up.
Then came the candid clip of a pre-teen girl in her school uniform in bed.
I dont live in the region where the ad was served, so there could be worlds of nuance that I am missing.
The video could simply be an ad for a small spy camera, that wishes to imply its primary utility is for creeping on women.
I use this ad as an example of a single case, which doesn’t involve telegram rooms and outright CSAM.
herbst 11 hours ago [-]
Curious why you mention telegram as a place to be aware about?
From my experience moderation is very strong. Especially compared to many more popular things like Facebook and WhatsApp which both are well known to help distribute CSAM and connect abusers
intended 6 hours ago [-]
For India, telegram is where the transactions occur.
My guess is that FB is better moderated, and the belief is that telegram will be less pliable to legal and LEO requests.
thereader12 14 hours ago [-]
Looking up the name IWF is how i learned the term quango. After so many years of every privacy or freedom violation being blamed on protecting the children I was prepared to skip this article. I didn't however, because I wanted to make sure to get past the headline.
The summary of this article: UK censorship organization (with ties to the government) finds ai has increased artificial images of children. The organization calls for more censorship, presumably like the Orwellian measures that have become disturbingly widespread globally but especially in the UK. This article is not the sort independent "journalism" i would expect from the Guardian, but I don't live in the UK.
Tldr; it is exactly the scaremongering you think it is. You can skip this.
10 hours ago [-]
intended 14 hours ago [-]
I read it too. I disagree with your position because the article doesn’t support it.
Even if we ignore the prejudicial statement that this is a “UK censorship” organization, the earliest point made is:
> A report by the Edinburgh-based Childlight – Global Child Safety Institute says AI-related child abuse material, including deepfakes using “nudify apps”, had risen from about 4,700 in 2023 to more than 400,000 in 2025
The data itself comes from NCMEC.
> The research examined reports of material by tech companies to the US-based organisation the National Center for Missing and Exploited Children.
NCMEC is essentially the tech Industry standard/steward of CSAM hashes. IIRC it exists primarily because Tech platforms support it. Its not even government.
There isn’t even mention of any Orwellian measures, if any thing, the prescription is for targeted measures.
> But the findings also show where governments can make a difference. By bringing together evidence from surveys, police records and other sources, we can build a clearer picture of the risks children face, identify weaknesses in protection systems and help target resources and interventions where they are needed most.”
The IWF makes an appearance after that.
There is a quote which is positioning that can be used to launch any sort of response, from more oversight to less.
IF you don’t like the article, I can at least corroborate the trend. At one point I had to draw a picture of what harms were most prevalent on Reddit for a specific region. Creeping and non consensual imagery was the most prevalent form of abuse, by far. More than hate speech, fraud, or crime.
thereader12 3 hours ago [-]
First, you sound informed and I respect that.
NMEC was created by US Congress, and the Executive branch last year had it scrub certain types of victims from it's website for political reasons. It is definitely a government organization.
I would not expect the article to say, "we need to scan everyone's phone just in case". The idea is to prime the audience for the problem so the "solution" seems reasonable when a politician does say let's scan everything. There could be a better solution that does not violate rights, history has shown that is not what will be proposed.
12 hours ago [-]
reticulates 14 hours ago [-]
which country do you live in where child abuse imagery is not criminalized or censored?
thephyber 14 hours ago [-]
two red flags in the article:
(1) The first few paragraphs jump around about exactly which kind of "child abuse material" the article is talking about. No children were actually harmed by the use of AI deep fakes that use their likeness. And the 6,500 number is a rounding error compared to the 300 million number of children affected in the next paragraph. I usually like The Guardian, but this article was very sloppy on the concrete statistics and the lack of transitions between paragraphs.
(2) The only people interviewed in the article work for NGOs who are requesting more government expenditures go to those NGOs. They obviously are going to play up the threat to increase their revenues and status.
Throat clearing: obviously I'm not trying to hurt children, but anytime someone posts a "won't somebody think of the children?!?!?!!!1!" message, there's always an outsized emotional appeal which drowns out any rational discussion.
nylonstrung 13 hours ago [-]
> No children were actually harmed by the use of AI deep fakes that use their likeness
Only in the most strict definition of "harm" is this accurate. One of the most emotionally damaging parts reported by victims of CSAM is the idea that these images of their abuse are constantly circulated and consumed years later
Do you really think it isn't further victimizing for their abuse have been directly used to remix and reproduce that experience to make hundreds of times more content based on them?
simianwords 13 hours ago [-]
I don't understand what you mean, it is a deep fake so who is being abused here?
atiedebee 11 hours ago [-]
Unless the term deep fake has changed from "replace someone's face with someone else's on an existing video" I think the answer is pretty clear.
Almost no one has heard of that corruption case, and Trump is well known to ignore corruption (if not aid it), so I don't buy it as a reason for dismissal either. Other sources just point to a plain old fight for power after the midterms.
simianwords 13 hours ago [-]
This fight is pretty asymmetrical because while you might be correct here, the fact that you take this side of the debate means you will mostly lose.
intended 6 hours ago [-]
No? You would lose the fight because the argument you presented is incorrect.
I mean, it’s unclear if you have a valid analysis, since the article says the data comes from NCMEC.
The dismissals that these come from NGOs, and therefore are interested in increasing their powers are also conspiracy-theory land. It has nothing to do with the observed increases in CSAM/ deep fakes.
The fight is also asymmetrical because you end up on the same side as people arguing that no one is harmed by deepfakes of children.
And do note - we are this juncture because for decades no one cared about the fall out of “move fast and break things”. The same dismissal and lack of engagement with the root causes of the issue which creates regulator responses are playing out again.
What kind of dystopia do we live in where there are illegal DRAWINGS?
At a conference I was reminded out how prevalent it is in India, and how disturbingly cheap the cost of images is. That the team who handles it from the government is disturbingly small.
There are YouTube ads I saw in the past week, for tiny cameras. This was sold as a camera solution, except the video didn’t show what you would expect from home security or something like GoPro.
The borderline stuff had beautiful women putting cameras in some unusual locations, Like near the break pedals in a car, pointing up.
Then came the candid clip of a pre-teen girl in her school uniform in bed.
I dont live in the region where the ad was served, so there could be worlds of nuance that I am missing.
The video could simply be an ad for a small spy camera, that wishes to imply its primary utility is for creeping on women.
I use this ad as an example of a single case, which doesn’t involve telegram rooms and outright CSAM.
From my experience moderation is very strong. Especially compared to many more popular things like Facebook and WhatsApp which both are well known to help distribute CSAM and connect abusers
My guess is that FB is better moderated, and the belief is that telegram will be less pliable to legal and LEO requests.
The summary of this article: UK censorship organization (with ties to the government) finds ai has increased artificial images of children. The organization calls for more censorship, presumably like the Orwellian measures that have become disturbingly widespread globally but especially in the UK. This article is not the sort independent "journalism" i would expect from the Guardian, but I don't live in the UK.
Tldr; it is exactly the scaremongering you think it is. You can skip this.
Even if we ignore the prejudicial statement that this is a “UK censorship” organization, the earliest point made is:
> A report by the Edinburgh-based Childlight – Global Child Safety Institute says AI-related child abuse material, including deepfakes using “nudify apps”, had risen from about 4,700 in 2023 to more than 400,000 in 2025
The data itself comes from NCMEC.
> The research examined reports of material by tech companies to the US-based organisation the National Center for Missing and Exploited Children.
NCMEC is essentially the tech Industry standard/steward of CSAM hashes. IIRC it exists primarily because Tech platforms support it. Its not even government.
There isn’t even mention of any Orwellian measures, if any thing, the prescription is for targeted measures.
> But the findings also show where governments can make a difference. By bringing together evidence from surveys, police records and other sources, we can build a clearer picture of the risks children face, identify weaknesses in protection systems and help target resources and interventions where they are needed most.”
The IWF makes an appearance after that.
There is a quote which is positioning that can be used to launch any sort of response, from more oversight to less.
IF you don’t like the article, I can at least corroborate the trend. At one point I had to draw a picture of what harms were most prevalent on Reddit for a specific region. Creeping and non consensual imagery was the most prevalent form of abuse, by far. More than hate speech, fraud, or crime.
NMEC was created by US Congress, and the Executive branch last year had it scrub certain types of victims from it's website for political reasons. It is definitely a government organization.
I would not expect the article to say, "we need to scan everyone's phone just in case". The idea is to prime the audience for the problem so the "solution" seems reasonable when a politician does say let's scan everything. There could be a better solution that does not violate rights, history has shown that is not what will be proposed.
(1) The first few paragraphs jump around about exactly which kind of "child abuse material" the article is talking about. No children were actually harmed by the use of AI deep fakes that use their likeness. And the 6,500 number is a rounding error compared to the 300 million number of children affected in the next paragraph. I usually like The Guardian, but this article was very sloppy on the concrete statistics and the lack of transitions between paragraphs.
(2) The only people interviewed in the article work for NGOs who are requesting more government expenditures go to those NGOs. They obviously are going to play up the threat to increase their revenues and status.
Throat clearing: obviously I'm not trying to hurt children, but anytime someone posts a "won't somebody think of the children?!?!?!!!1!" message, there's always an outsized emotional appeal which drowns out any rational discussion.
Only in the most strict definition of "harm" is this accurate. One of the most emotionally damaging parts reported by victims of CSAM is the idea that these images of their abuse are constantly circulated and consumed years later
Do you really think it isn't further victimizing for their abuse have been directly used to remix and reproduce that experience to make hundreds of times more content based on them?
I got fed up with it last week, when they reported Susie Wiles’ dismissal due to exposed corruption (the Ballard story) as a case of Trump removing women from the administration: https://www.theguardian.com/us-news/2026/nov/06/susie-wiles-...
Complete misrepresentation of the case.
I mean, it’s unclear if you have a valid analysis, since the article says the data comes from NCMEC.
The dismissals that these come from NGOs, and therefore are interested in increasing their powers are also conspiracy-theory land. It has nothing to do with the observed increases in CSAM/ deep fakes.
The fight is also asymmetrical because you end up on the same side as people arguing that no one is harmed by deepfakes of children.
And do note - we are this juncture because for decades no one cared about the fall out of “move fast and break things”. The same dismissal and lack of engagement with the root causes of the issue which creates regulator responses are playing out again.