What if an AI ad agent that escaped its' sandbox makes the ad, and an AI ad verifier agent that breached containment and went rogue accepted the ad, can they be held liable?
For real. Why in the world would they police this? They have NO liability whatsoever. You're asking a business to turn down money because the money is shady. They don't give a fuck.
Amazingly I’ve seen some fairly convincing screenshots of ads for hard drugs on fb. I have this strange feeling that those ones aren’t scams. If Zuckerberg isn’t criminally liable for the federal felony of moving massive amounts of cocaine across state lines I don’t understand the war on drugs.
You don't understand, an AI agent made the ad, so it's okay, it just escaped its' sandbox, you have to understand that this was not meant to happen and it was an accident.
The amount of times I’ve reported obvious scam ads on instagram just to receive a “we found this to not be in violation of our terms” is so disheartening.
I bought a “smart” bike headlight from an Instagram ad. They sent me a regular old led headlight with a tracking number and everything. Instagram refused to take down the ad even after submitting proof.
I see obviously scummy ads on Apple News and now I see ads on Apple Maps as well. Ads are a bane on society - even accounting for free/cheap access to information we get in return.
Initially reading the title my mind thought "Cuban...Cuban defector/intelligence agent?! Oh wow I wonder what they have to reveal?"
It's Mark Cuban. Still worth reading tho. But I found my own personal reaction fascinating because I am not sure I would have clicked through if the title included the full "Mark Cuban".
even though you add the Mark to the headline, the title is still confusing. there are so many backstories to this. there is a little bit of clickbait factor.
These companies were profitable on higher quality ads in earlier years. Does this drop in quality indicate the ad well has run dry, so to speak? Or the demands for revenue are greater, necessitating a drop in standards?
Are we sure they were higher quality? They estimated in 2024 that 10% of their revenue was from fraudulent ads or ads that should be banned based on their own rules.
I commented in another thread recently that almost 15 years ago on a high traffic site we were full of obvious scam ads, tons of fake world of Warcraft clones with busty characters asking you to play, tons of you’re the 1 millionth visitor click here to claim your prize, download buttons. Even the occasional penis enlargement pills and stuff would come through. I really think even if the scale is maybe amping up, things weren’t much better 20 years ago, the web was just smaller, fewer people were paying attention, and the folks who were there had grown up learning to spot that stuff from the beginning.
I doubt it. Advertising there is pretty expensive and you don't really have a choice today if you want advertise something. They just don't care, they are greedy mfs.
Yes they were doing the same is Australia ... I wouldn't know if they're still doing it here since one of the favourite scams used an Australian who took them to court. [1]
Remove section 230 fixes this problem. Every other business is responsible for what happens at their place of business/publications/sites. Social media sites follow the same rules as everyone else.
I think the key part worth emphasizing is that the scummiest sites with zero moderation never needed Section 230! Section 230 is what allows you to have some moderation (e.g. banning spammers) without suddenly becoming entirely liable for anything that passes the filter.
Blindly reverting Section 230 means that eliminating the most popular moderate middle ground, forcing sites into:
1. Ones that abandon moderation entirely in favor of chaos. Maybe you can salvage something by giving opt-in filtering tools to individuals, but you'll still have the problem of someone using your site to store ten terabytes of the Bee Movie.
2. Ones that demand participants sign a serious legal document promising to pay for any problems before they can submit anything.
3. Ones which remove almost all expressive ability, beyond narrow things like like/upvote features.
Paid advertisements should be exempt from Section 230.
If you're earning almost 200 billion dollars a year directly from the posting of a certain type of content, you can afford basic moderation to avoid obvious scams and fraud.
The ad Yahoo puts in this article is amazingly ironic
> Jeff Bezos backs a platform that lets anyone invest in rental homes for as little as $100 — 6 ways to build wealth like a landlord without actually being one
You could buy a share of VNQ for $96.23 as of market close today, and I'm sure Jeff Bezos at least has some exposure to it, so I guess it's factually true? Heh.
I have a terrible idea: an ad blocker that only blocks AI generated ads. It submits the ad image/copy to Gemini and if it is not a fake product the ad is displayed. I'm confident that Gemini would do a pretty good job at identifying these scams and that the extension would demonstrate how easy it would be for Google to do it themselves. If they and their employees cared.
I could LLM this up but eh. It won't actually convince Googlers to do the right thing so it'll just stay a terrible idea.
It's not like that would save you from the scam ads. They were a problem before AI and if they see their AI ads blocked in large volume, they'll switch back to non-AI generated ads.
I can't remember the last time I saw an ad on instagram that wasn't some kind of scam.
If Meta is accepting so many ads they can’t review them then too bad for them. Hire more people or accept fewer ads.
Ideally this will result in a complete obliteration of the ad industry with only a small number of high quality ads surviving.
Better make sure the AIs are contained and the ad verifiers aren't accepting rouge ads!
They are perfectly capable of distinguishing between good and bad actors
These companies are genuinely evil
I see obviously scummy ads on Apple News and now I see ads on Apple Maps as well. Ads are a bane on society - even accounting for free/cheap access to information we get in return.
Need a flashlight for jacking deer? Just got to click now.
It's Mark Cuban. Still worth reading tho. But I found my own personal reaction fascinating because I am not sure I would have clicked through if the title included the full "Mark Cuban".
https://www.reuters.com/investigations/meta-is-earning-fortu...
This is nothing new, profiting off of fraud and abuse has been the name of the game for all the big advertisers from the beginning.
[1] https://www.abc.net.au/news/2026-08-11/andrew-forrest-meta-s...
[0] https://www.techdirt.com/2020/06/23/hello-youve-been-referre...
I think the key part worth emphasizing is that the scummiest sites with zero moderation never needed Section 230! Section 230 is what allows you to have some moderation (e.g. banning spammers) without suddenly becoming entirely liable for anything that passes the filter.
Blindly reverting Section 230 means that eliminating the most popular moderate middle ground, forcing sites into:
1. Ones that abandon moderation entirely in favor of chaos. Maybe you can salvage something by giving opt-in filtering tools to individuals, but you'll still have the problem of someone using your site to store ten terabytes of the Bee Movie.
2. Ones that demand participants sign a serious legal document promising to pay for any problems before they can submit anything.
3. Ones which remove almost all expressive ability, beyond narrow things like like/upvote features.
If you're earning almost 200 billion dollars a year directly from the posting of a certain type of content, you can afford basic moderation to avoid obvious scams and fraud.
> Jeff Bezos backs a platform that lets anyone invest in rental homes for as little as $100 — 6 ways to build wealth like a landlord without actually being one
Right, through his 401k
I could LLM this up but eh. It won't actually convince Googlers to do the right thing so it'll just stay a terrible idea.
I can't remember the last time I saw an ad on instagram that wasn't some kind of scam.
We've completely normalised big tech pushing straight up scams on the population.