Hacker News Reader

Comment by stusmall | Hacker News Reader

Once I got an Instagram ad for buying ketamine that just linked to a telegram channel. They didn't even bother being coy or using mispelling or slang. A simple keyword search to flag for more review would have caught it. I can't even wrap my head around what internal controls exist when something like that makes it out to users.

View on HN

The bad actor serves a benign ad to the ad review system, and only serves the scam to real users. It's called "cloaking" - an interesting (but a bit depressing) topic to explore.

So the ad review system is just requesting the ad from the advertiser, and not ever bothering to disguise itself? Didn't we have this shit figured out for brick-and-mortar restaurant reviews decades ago?

As far as I understood, the problem happens when the ad has a link to a website. I can't imagine that happening with static images or videos that don't link a website (that could be solved quite trivially...)

You can change the link target, but you cannot change the media afterwards without a new check.

That's called a broken system and implies either incompetence or malicious intent by the ad platform.

The fact the ad platforms are multi-billion dollar companies leans my opinion towards malicious intent.

There are Facebook groups devoted to selling stolen cars (strikers).

There’s zero to none manual review. The people who run these type of ads probably burn 100 facebook ad accounts per day

Oh definitely. I have zero expectation of high level of manual reviews. You can run limited runs of adds for next to a couple dollars. The math could never work out. I understand a lot with make it through the system, this was just so blatant. It should be so easy to catch with an automated system. It was nothing but red flags. The automated systems could reject outright or maybe escalate to manual review if it met enough criteria (account reputation, spend floor, etc)

>The math could never work out.

Horseshit. Running an ad in a local publication was also pretty damn cheap and was always human reviewed.

Meta wants to have big scammers. It's hugely profitable. That's why they codified internal policy that if the scam generates at least 0.15% of Meta's revenue, they must be protected and never moderated.

There was a big "bombshell" report on this yesterday (it didn't hit HN frontpage though).

Meta knowingly gets a lot of its funding from scams. They love it! They don't care about technical solutions for it: they've banned any solutions from being implemented because it would impact revenue.

Zuck and Meta are playing the role of Robinhood. They take the money from scammers and advance their philanthrophy.

https://x.com/a16z/status/1986486508355002584

Priscilla Chan and Mark Zuckerberg on Curing All Disease

We sat down with Dr. Priscilla Chan and Mark Zuckerberg, co-founders of the Chan Zuckerberg Initiative, to discuss their ambitious plan to cure, prevent, and manage all disease by the end of the century.

I’m assuming this is satire but in case it’s not- He’s PROVIDING A SERVICE for the scammers. Not stealing from them lol

Did Facebook's LLM write this comment?

Zuck has donated something like 2% of his net worth. Not really “ambitious”. For a normal American that’s like donating to “cars for kids” every 8 years when you upgrade your Mercedes and claiming you’re a philanthropist

Reply on news.ycombinator.com