Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Meta Ran Ads That Contained AI-Generated Child Sexual Abuse Imagery

wired.com

Post reply on HN
Print view View original

171–180 of 294 posts

Re: Meta Ran Ads That Contained AI-Generated Child Sexual Abuse Imagery

#171

Post by shakna »

newAccount2025 wrote: post #168
7952 wrote: post #100

Earlier quoted context omitted.

They have the resources to hire a vast number of people. So yes they probably could manually review every ad. And either way it is their problem to solve. They are not entitled to be hugely profitable or free from the obligations that smaller organisations have.

Surely there is no need to review every ad. Coke and Apple and whatnot are not going to be doing this. New companies or companies with a history of complaints can face extra scrutiny. With AI capabilities so far along you have tier 1 ready for free…

Every advert for TV and radio has to pass a review in most jurisdictions. Billboard laws vary wildly, but some also require a review.

Why not ads online? The industry would adapt pretty much overnight, because it is something everyone already does.

Re: Meta Ran Ads That Contained AI-Generated Child Sexual Abuse Imagery

#172

Post by nitwit005 »

craftkiller wrote: post #60

I've noticed on YouTube I see a lot of adult sexual ads. Things like home cleaning ads that have an unmistakable wood-grain image of fellatio. I've always wondered "how are these getting past the moderators?". I guess the answer is the obvious one: no one is moderating anything.

The more surprising thing with YouTube is how much obvious copyright infringement is in ads. The whole platform seems oriented around copyright enforcement, but ads seem to get a pass.

Re: Meta Ran Ads That Contained AI-Generated Child Sexual Abuse Imagery

#173

Post by karakoram »

craftkiller wrote: post #60

I've noticed on YouTube I see a lot of adult sexual ads. Things like home cleaning ads that have an unmistakable wood-grain image of fellatio. I've always wondered "how are these getting past the moderators?". I guess the answer is the obvious one: no one is moderating anything.

The worst ads are the shady games with all sorts of sexual references. They don't end no matter how many times you report it, block the ads or select the "less gaming ads" option.

Re: Meta Ran Ads That Contained AI-Generated Child Sexual Abuse Imagery

#174

Post by jlarocco »

craftkiller wrote: post #60

I've noticed on YouTube I see a lot of adult sexual ads. Things like home cleaning ads that have an unmistakable wood-grain image of fellatio. I've always wondered "how are these getting past the moderators?". I guess the answer is the obvious one: no one is moderating anything.

Is it surprising? Google and Facebook can barely police the content their human users post.

Re: Meta Ran Ads That Contained AI-Generated Child Sexual Abuse Imagery

#175

Post by Ajedi32 »

mplewis wrote: post #78
Ajedi32 wrote: post #71

Earlier quoted context omitted.

As it should be. If the cost of fixing the problem is greater than the amount of harm the problem is actually causing to society, then it shouldn't be fixed. Set fine amounts accordingly. (Whether we actually do set fines this way is, of course, questionable. And probably this particular problem would get fixed even if there were no fine merely because of the reputational damage it's causing, but I'm speaking in gene…

you don't think one of the world's biggest media platforms disseminating CSAM is causing meaningful harm to society?

Yes it's meaningful, and the fine amount should reflect that. But hypothetically, if it's (just throwing a random number out there) 1 image in 50 billion, and Facebook gets more benefit from the 50 billion good images than society loses from the 1 bad one, and society recovers its loss from Facebook via a fine, but the fine isn't big enough to make them change their behavior, how exactly is that a problem?

Re: Meta Ran Ads That Contained AI-Generated Child Sexual Abuse Imagery

#176

Post by alex1138 »

burningChrome wrote: post #67
CrzyLngPwd wrote: post #11

All their fines are merely a cost of doing business, and they won't change their ways until the fines hurt.

This is so true. When you have a billion users, you can literally lose hundreds of millions of users and it wouldn't make a dent. The only way is to get the government involved and regulate this and just keep hammering them with massive fines that really make a dent in their bottom line. For a lot of people, government regulation of this sort is not desirable, but at what point does someone need to drop the hammer to…

> When you have a billion users

What if they make a shadow profile for you? Do you count as a user? What if you've tried to delete your account and you can't? (I meant a real account here, not even a shadow profile. People have had the things reactivated years later)

Re: Meta Ran Ads That Contained AI-Generated Child Sexual Abuse Imagery

#177

Post by tbrownaw »

randyrand wrote: post #159

Non-specific terms are a pet peeve of mine in accusations. Is this new term broader than the old term of CP or is it the same? Why did we even switch?

> Is this new term broader than the old term of CP or is it the same? Why did we even switch?

The purpose of the new term was to emphasize that producing it involves the actual abuse of an actual child.

And then somehow it got applied to drawings and lately computer-generated stuff.

Re: Meta Ran Ads That Contained AI-Generated Child Sexual Abuse Imagery

#178

Post by nitwit005 »

meric_ wrote: post #57

I'm wondering what people expect these companies to do? Human moderators? Sure, but Meta products have approximately half the worlds population using them. At this scale it's pretty mind bogglingly difficult to moderate down everything. The ads/posts served per second is an astronomical number. This truly is a case where AI is the solution. There's no way to automate this at scale with humans. As AI gets better hopef…

The ads make money though. The cost to approve the media use and have someone quickly glance at it is going to be tiny compared to their revenue.

Re: Meta Ran Ads That Contained AI-Generated Child Sexual Abuse Imagery

#179

Post by duxup »

craftkiller wrote: post #60

I've noticed on YouTube I see a lot of adult sexual ads. Things like home cleaning ads that have an unmistakable wood-grain image of fellatio. I've always wondered "how are these getting past the moderators?". I guess the answer is the obvious one: no one is moderating anything.

Man I got a “clothing brand” last night on YouTube of a VERY young girl wearing… not much, and it didn’t fit well in all the wrong / revealing places…

Reported the add but damn YouTube, the freaky fake medical ads are bad enough.

Post reply on HN