54 comments

[ 2.0 ms ] story [ 115 ms ] thread
On one hand it's not great to censor but on the other it's their platform and they would like to protect it.
They shouldn't be underhanded about it then, misrepresenting as phishing just make them even more spineless.

If they had been clear about it, posting a warning "we consider this post detrimental to our business" go ahead and censor it, being underhanded just made it stink much, much more.

They won't do that because the path they're taking now has the "air" of a process run amok, that wasn't "intended" to be used for this purpose. If they outright said "we're going to censor content that is critical of Facebook" there would be such a massive outcry that it could potentially mark the end of Facebook.
They have a right to protect their platform, and we have a responsibility to criticize them publicly when they lie about it or do it in a harmful way.
I have seen ads on facebook that are phising links. lol
exactly. have you ever tried to watch youtube without adblock? I started reporting them when they temporarily were able to get around adblock because they were obvious scams, but thats 90% of the ads on there. google/facebook ads are barely above pornsite advertisements as of 2024.

the only people clicking on them are extremely low info olds

My funniest instance was reporting an obvious crypto scam as potential SPAM or fraud to LinkedIn. It featured Wayne Gretzky promoting a can't-miss investment of lifetime.

It took less than 10 minutes for the Safety and Security team to thank me for the report, but that the ad was determined to be legit

Legit means the payment cleared.
Wayne Gretsky is on MGM gambling ads on TV. Putting us image in crypto scams isn't any worse than his "legit" ads.
Seems a little different, because MGM admits it's gambling, while the other claims to be an investment
Those ads on facebook for some random dating site, i had stumbled upon 10-15 ads in few hours while scrolling reels. I reported each one of them, until I got tired and uninstalled facebook ffs.

The weird thing about some of those ads were full NSFW content. Literally Porn. During reporitng the ads, it says "It was targeted to People from the age of 19 - 30, living in country <Some Asian Country>. It was a third world country. It seems like they don't fukin care who promotes what contents in such countries.

This is the entire text of the linked Bluesky post (which is part of several threads):

> Looks like it a few domains, including thehandbasket.co, were mistakenly classified as a phishing site, which has since been corrected as Andy mentioned. Unfortunately, at our scale, we get false positives on safety measures all the time. Apologies for the trouble.

False positive that was already corrected at the time is was posted on Bluesky and corrected long before it was posted here on HN.

This all happened and was fixed days ago, but people are still trying to post headlines that imply something more nefarious is still happening.

There is no story here, just journalists trying to stir up controversy and imply malice. Unfortunately this is a valid tactic for driving attention to small substacks and similar outlets who wouldn’t otherwise be noteworthy. Getting “banned” by Facebook is the most effective thing they could ask for to gain more exposure, as evidenced by the way this has shot to number 1 on Hacker News despite having been corrected days ago.

> Unfortunately, at our scale, we get false positives on safety measures all the time.

At this point, the “at our scale …” argument is such an arrogant excuse for doing damage all over the world. If I can’t ensure people are treated fairly, don’t get harassed, etc, maybe it’s not a viable business then. Stop offloading all kinds of externalities onto the public.

The "doing damage all over the world" framing makes no sense here. Blocking a link on your own website is not an externality by definition.
So, Facebook allowing content is bad, and Facebook not allowing content is also bad?
Well, yeah. No entity should be in the centrally authoritative position that Meta alongside a handful of other companies is.

Moderation and filtering needs to be decoupled from content hosting and authn/authz. This is a major part making me optimistic about bluesky/atproto.

Are you implying that would be some kind of fallacy?
In most cases I would agree, but in this case they actually did do reputational damage. There are people asking the site owners why the site is considered insecure.
I didn't consider that aspect. I admit that does sound frustrating from the perspective of the site owner.
What are they supposed to do? Not have any automated filtering? Just shut down when the site gets too big.

I'm not a fan of these privately owned social media platforms with no oversight, fairness, or recourse for the public, but people want corporate social media and I'm not sure what these giants are supposed to do or what people even want here.

> What are they supposed to do? Not have any automated filtering? Just shut down when the site gets too big[?]

Maybe? It's not that crazy of an idea that if you can't serve your clients that you are put out of business?

But they do serve their clients. We don't shut down hospitals because one doctor misses a diagnosis.
The hospital is serving a public good.
> hospital is serving a public good

Given hospitals are frequently privately owned, one could argue Facebook provides a public service in running a social media platform. Discriminating models based on public goodness is simply command-economy picking of winners and losers based on politicians’ limited knowledge and preferences.

One doctor missing a diagnosis doesn’t indicate a systemic issue with the hospital. A hospital which drove teenage girls into depression and suicide via an addictive product would rightfully be shut down.
Well let the magic hand of the market decide! I'm sure people will be flocking out any time now and advertising revenue will dry up!
The best (but not necessarily good) answer I see is that ad supported social media can’t meet the minimum bar for effective moderation and be profitable, then it should be shut down as a nuisance. Not all business models have a right to exist.

Perhaps if free/ad was not workable we could see subscription services with better moderation. If not (people won’t pay), then it’s not actually worth that much to users so it’s fine to let it die.

Well said. I would add another option, they should be public utilities
> What are they supposed to do? Not have any automated filtering?

Just give the user a warning that Facebook sees some likelihood that the site is a phishing site, and let the user decide by himself whether he wants to follow the link or not.

What sounds a over man show do with no revenue?
If we presuppose that platforms should be able to scale past their ability to not cause harm or even comply with the law, this defense makes sense. It’s not feasible for a company with a billion users to avoid harming people. This is the philosophy we’ve been living under for at least the last 20 years.

If we instead say that people are more important than a corporation’s right to scale, then a single large-scale violation would have been the end of Facebook and have criminal consequences for the leadership of the company. The scale of the act should make you more culpable, not less.

This piece of “journalism” isn’t fair either though. Should we shut down CNN too?
How many people would follow links to that website if Facebook didn't exist? Fewer than of Facebook existed as accidentally banned them for one day.
We can't know since history has no ifs and we have no idea what else could have been in place of Facebook.
At that scale, they have the resources to deal with it.
Can you say "clueless"? I didn't think so.

What you're saying is that if something isn't perfect, it shouldn't exist at all. And guess what: Nothing is perfect.

Wrong, first they blocked the Kansas Reflector but when they fixed that the “offending” column was still blocked. That column was reposted by someone else and their whole website was also blocked.

There is controversy and malice

The quote in my post was verbatim copy and paste of the linked BlueSky post.

If you have different information, perhaps post a valid source here rather than saying “wrong” and then injecting something else. I literally just copy and pasted the linked post text.

The linked post is halfway through a BlueSky thread that has a screenshot of a part of a Twitter thread about a link to something else that I can’t click. Nothing screams “social media ragebait mill” like a long chain of unclickable screenshots to some ragebait take about some story that happened elsewhere.

I think GP may have been responding to this

> There is no story here, just journalists trying to stir up controversy and imply malice.

Which was not a quote

Assuming you’re not familiar with this story at all, here is some previous discussion of the numerous events in which the critical article was stopped from being shared on meta’s platform. https://news.ycombinator.com/item?id=39945811

Tell me, how does the same story, reposted in three different places, conveying criticism about the source of the blocking, lead to phishing issues on three different domains?

It is clearly underplaying questionable behavior that has caused serious problems for a media outlet that were only fixed because this got attention.

Obviously because something in the content triggered the phishing filter.
And what’s in the content that would trigger phishing? Among what you’ll find in the piece:

- Criticism of Facebook’s ad policies and functions

- Discussion of climate change issues

- Links to Vox, government sources, and a bunch of PBS sites

https://kansasreflector.com/2024/04/04/when-facebook-fails-l...

This gets blocked, yet I see significantly spammier links on Facebook by accident. To me, the real issue is likely that their filters aren’t very good.

Or the problem is far more complex at their scale than you are considering, and ie simple content filtering is but a mere tiny signal amongst what qualifies links to be filtered.
You're not making an argument that makes Facebook look non-malicious here.
so that's not the case - the Facebook guy was lying. At absolute best, he was making a much stronger claim than he confidently could.

There's continued to be intermittent bans on the Kansas Reflector site, and there's even fresh bans on blog posts talking about the ban on the site. Also, the original KR post about Facebook still seems to be blocked from being posted on Facebook or Instagram. Someone is actively adding these URLs to the filter.

There's this weird tendency to bend over backwards to assume Facebook, a company repeatedly convicted of doing malicious things, wouldn't do a malicious thing, and none of the excuses stand up to the facts for a second. Would a company run by a man who openly compares himself to Caesar Augustus try to silence its critics?

Irony is that you post this on hackernews that is known to censor critical things that could affect ycombinator negatively. If you use a centralized service be prepared to get censored by the providing overlord if it goes against his agenda.
(comment deleted)