NY Orders Apps To Lie About Social Media ‘Addiction,’ Will Lose In Court
from the compelled-speech-101 dept
New York Governor Kathy Hochul just signed a law that’s going to get expensive fast. The state’s new social media labeling requirement—S4505, courtesy of state Senator Andrew Gounardes—forces websites to slap unscientific warnings on their services claiming that features like algorithmic feeds and push notifications cause addiction.
It’s compelled speech based on contested science, which means it’s blatantly unconstitutional, and courts have already rejected nearly identical schemes. But Hochul and Gounardes either don’t care or don’t understand First Amendment basics, so New York taxpayers are about to fund a losing legal battle whose only real purpose is generating headlines for those politicians pretending to care about “protecting the children.”
(For those keeping track: Gounardes has a track record of pushing unconstitutional anti-internet legislation, and when constituents call him out on it, he tells them they’re just parroting “big tech” talking points. Several of his constituents have forwarded me these dismissive responses. Engaging on substance apparently isn’t part of his process.)
We had warned New York officials that this law was unconstitutional, but Hochul and Gounardes clearly don’t care.
The law targets any site with algorithmic feeds, push notifications, infinite scroll, like counts, or autoplay—basically every modern website—and demands they post warnings claiming these features are addictive. This isn’t based on scientific consensus. It’s based on misreading correlational studies and pretending they show causation.
Hell, even with Texas’s law regarding porn which went to the Supreme Court over its age verification requirements, every court along the way outright rejected the required labeling with unscientific warnings about how porn is unhealthy. Not just the lower court, but the Fifth Circuit, which seems quite willing in other circumstances to go along with anti-internet nonsense. While the Fifth Circuit’s ruling in the case was extremely problematic on other grounds, even it wouldn’t go so far as to say that the government can force websites to label their services based on made up, unsubstantiated claims of “public health.”
Throughout that ruling, the normally censorial Fifth Circuit highlights that the consensus on such public health warnings for pornography are way too contested to require getting around the First Amendment’s prohibition on compelled speech around health warnings:
We are not scientific journal editors, much less social scientists, behavioral experts, or neurologists. The courts generally are not the place to hash out scientific debate, particularly not on so contentious a topic as the impacts of engaging with pornography. Experts must do that in academic journals, studies, and presentations. Therefore, the record leaves us with no option but to declare that the health impacts of pornography are currently too contentious and controversial to receive Zauderer scrutiny.
The science on social media’s “addictive” features is even more contested than the porn stuff the Fifth Circuit already rejected. Yet Hochul’s signing statement declares, with absolute certainty, that “studies show” spending more time on social media increases anxiety and depression. That’s likely backwards.
The studies show correlation, not causation. And actual experts in the field—not politicians cherry-picking abstracts—suggest the causal arrow likely points the other direction. Kids struggling with anxiety and depression aren’t getting adequate professional help and support. So they turn to social media to self-medicate, to find community, to cope. The increased screen time is a symptom, not the cause. Which means taking away that tool—or scaring them away from it with dire warnings—could make things considerably worse for the kids who need support most.
But the key here is that there is no generally agreed upon scientific consensus that social media itself, let alone any specific features of social media, are the cause of depression or anxiety. Yet, Hochul and Gounardes are forcing all sorts of sites to speak and claim that these unproven things are true. That’s dangerous! And obviously unconstitutional.
The bill’s language is so broad that it’ll snag far more than just Facebook and TikTok. Here’s the language in the bill:
“Addictive social media platform” shall mean a website, online service, online application, or mobile application that primarily serves as a medium for covered users to interact with media generated by other users and which offers or provides covered users an addictive feed, push notifications, autoplay, infinite scroll, and/or like counts as a significant part of the services provided by such website, online service, online application, or mobile application.
Techdirt has our comment voting system, which we created years before other systems of “likes” and whatnot, but do we need to warn users that our comments are addictive? Here’s the definition of “like counts”:
“Like counts” shall mean the quantification and public display of positive votes, such as but not limited to those expressed via a heart or thumbs-up icon, attached to a piece of media generated by a covered user.
We don’t publicly “quantify” people voting for “funny” or “insightful,” but we will display an icon when they reach a threshold. Does that qualify? Do I need to pay a lawyer to find out if I have to warn people our comments are addictive? Can I send the bill to Andrew Gounardes?
The whole thing is nonsense.
There can be exceptions, but they appear to be solely the determination of NY’s Attorney General deciding which sites to exempt, which seems to be putting way too much power in the hands of a single politician.
“Addictive social media platform” shall not include any such service or application which the attorney general determines offers the features described herein for a valid purpose unrelated to prolonging use of such platform.
We’ve criticized NY’s Attorney General in the past. Now she has a tool to punish us since she has sole authority of determining whether or not our comments qualify under this law. Doesn’t that seem like a problem? It does to me!
Obviously, this applies to way more than just Facebook and Tiktok. But also more than just our comments. It means almost any site with a recommendation algorithm? That’s probably covered. Auto-playing videos? Covered. Notifications of things that may interest you? Covered. That means news sites, recipe apps, fitness trackers, even email clients could theoretically fall under this law’s scope if they have the wrong combination of features and don’t play nice with NY’s Attorney General.
All of these services could face fines from the AG—the same AG with a documented history of launching bogus investigations against internet companies specifically to cover up New York’s own policy failures.
So what happens next? Either this law gets struck down immediately when challenged in court—wasting taxpayer money on litigation over compelled speech that even the Fifth Circuit wouldn’t tolerate—or it survives long enough to force websites to lie to users about nonexistent dangers, potentially driving vulnerable kids away from the online communities that actually help them cope.
Neither is a good look for New York. But at least Gounardes and Hochul will have gotten their headlines. That must be worth something.
Filed Under: 1st amendment, addiction, addictive feeds, algorithmic feeds, andrew gounardes, compelled speech, infinite scroll, kathy hochul, likes, new york, notifications, social media, social media addiction, warning labels, zauderer
Comments on “NY Orders Apps To Lie About Social Media ‘Addiction,’ Will Lose In Court”
Techdirt is way too addictive to my taste.
See you in court Mike.
'The truth doesn't support my claim so I made a law to force you to lie for me.'
A law that required politicians and would-be politicians to publicly post warnings in their press releases and on their campaign websites that politicians are known to lie and mislead people, accept money in exchange for preferential treatment of businesses, industries and individuals, and go into politics for their own gains rather than to ‘serve the public’ would be far more accurate so by their own logic they should all be legally required to post all of that, even if they don’t agree that it’s true.
Ahh, I see you found the actual addiction at play here.
By the way, New York State Senator Andrew Gounardes maintains an active presence on X, the Nazi CSAM app. So does NY Attorney General Letitia James. And NY Governor Kathy Hochul. Maybe they should put their money where their mouths are when it comes to social media?
Re:
Sadly politicians in the US are bought and paid for by the CEO’s that own the Nazi CSAM platform/apps (Facebook, Twitter/X).
Indeed.
NY lawmakers… addicted to misinformation.
May contain nutjobs
Here’s a thought. Treat it like all those food products with “may contain traces of nuts” warnings on the labels.
If Techdirt’s servers decide the user is in NY then shove up a big ugly splash screen with “may contain addictive features such as [the complete list]” with three buttons: “close the window and leave the site”, “ignore and continue to the site”, and “inform yourself on this topic” which leads to links of the various excellent TechDirt articles debunking this nonsense.
Perhaps also include a “write your governor” button leading to an email window pre-addressed to “Dear Kathy” plus some sample paras that can be dragged-and-dropped. You could even add a veneer of “balance” by providing paras containing both pro (drawn from Kathy’s own press releases) and con stances (the science). It’d be interesting to see which paras were chosen by NYers because, in my experience, they don’t take kindly to BS.
Re:
This is what any site should do.
Most are only correlational, but a handful look at causation. e.g. Braghieri,Levy, Makarin (2022) was causal. It’s not enough to establish a consensus (and yes, some of them even find null or positive effects), but not all the studies are correlational.
Re:
There is generally sparse causational study, and absolutely nothing approaching a consensus or even a serious “this is interesting” on causation, but thanks for the pointless equivalence pedantry.
Re: Re:
So then why does Mike keep saying there is none? Not that it’s sparse, or weak, or unconvincing. Very specifically, does not exist.
I specifically said they weren’t equivalent/no consensus. But I don’t think it’s pedantry to point out it exists and that it’s being misleadingly summarized. It’s relevant, important data, which is why it’s being avoided. Pretending it doesn’t exist misrepresents the literature, and it’s not pedantry to point out cherrypicking.
There definitely is an interest in it. There is a consensus that more causal data would be useful, and it’s consistently called for in the literature (preferably while distinguishing heterogenous groups), including in the links Mike is happy to cite as proof. The reason the data is sparse is because causal studies are hard to do, not because of lack of interest. You need something like a natural experiment like in Makarin.
And this is exactly why I point it out. You’re trying to paint it as if it’s a settled issue, when in reality there are still big open questions that the field wants to answer. Being so unequivocal about it misleads people who won’t go looking at the literature. It’s an unjustifiable, intentional, and easily avoidable choice. Which is especially disappointing because you can make a perfectly reasonable argument without doing that. There’s no excuse for it.
There’s an interesting quirk in the definition of (ugh) “Addictive social media platform” as well as the definition of “Addictive feed” this bill borrowed from the SAFE Act. Neither one applies unless these features are used to serve “media generated by other users.” Most blogs aren’t going to be covered by the law, though that wouldn’t stop an AG from harassing bloggers – this is basically a top-down SLAPP law, after all.
I say this as a reminder that this has always been about user-generated content. If the problem is that infinite feeds or algorithmic recommendation are “addictive”, it shouldn’t matter who made those things that are displayed. But this isn’t about “addiction,” it’s about elite media types who are sick of the peasants stinking up their drawing rooms and politicians who hate being mocked by their lessers.
But if you want to do these things safely, just do them in a podcast. Thanks to bad drafting, the SAFE Act (which, again, provides all of the definitions) defines “media” as “text, an image, or a video.” Audio is exempt – you can recommend that all day.
Yeah idk. Facebook absolutely thinks these features are addictive, and they tune them to be as addictive as possible. Social media platforms are engineered to commandeer your attention, to keep you engaged even if that means giving you harmful/toxic/enraging content.
This bill might be wrong-headed based on its wording and lack of scientific consensus on the addictiveness of specific features, but i have absolutely no problem with states trying to rein in the abuses of companies like X and Facebook who absolutely, in my mind, abuse their users.
Re:
You’re making unsubstantiated claims; and as much as i hate corporate (social) media, i don’t see the government here doing anything at all to rein in any abuse. It’s bloody unconstitutional theater, full stop.
Re: Re:
There’s been leaks that Facebook has internal data showing they know the site is harmful for its users.
Addictive? In a literal medical sense? I suppose not.
But it is harmful, the companies know it, and they continue, deliberately, to operate in such fashion as harms their users. Warning people about this harm is no problem in my mind.
If cigarettes still caused cancer but weren’t physically addictive, I’d still support putting warning labels on them. Facebook doesn’t need to be addictive to be harmful, Facebook knows their platform harms their users, so it makes sense to warn them.
And we may still find more evidence of causation. But I’ll concede calling it addictive without consensus on that specifically is misguided.
Re: Re: Re:
That’s misleading in the extreme. There were leaks of internal research, which was later released publicly, showing that Facebook studied 24 different traits to see if their service was seen (in user surveys) to make people feel worse about themselves. They found it was the case in only one of the 24 (for teen girls & body image). And the reason they did that research and called it out in an internal report was to seek to FIX that.
But then the report was leaked to the media and the media ran with “FACEBOOK HAS INTERNAL DATA KNOWING IT’S HARMFUL TO TEENS!” effectively guaranteeing that no website will EVER do such internal research to try to fix problems again.
And you’re contributing to that.
This is just wrong. Again, most companies were looking at ways to reduce the harm, but people like you turned that into “they know it and they’re doing nothing about it” which then makes them not even want to look at ways to reduce harm because people like you will claim that their efforts are because they don’t want to reduce harm.
You’re a part of the problem.
Actual experts say the opposite. Candice Odgers, who literally studies this stuff, has called out how damaging it is to kids to tell them things that most people know how to use well and in a useful manner are inherently harmful.
Again, you are a part of the problem.
It’s great to see that Mike is still sticking up for the biggest tech companies. I agree with the law, these people should have to put warnings on their sites. Just like cigarettes or dangerous chemicals. Sorry to have to put it this way, but the products are obviously designed to be toxic, anyone who suggests otherwise is not being realistic.
Re:
Like literally half the point of the article is that this will do massive harm to SMALL companies, not the large ones. All this does is help the biggest companies by making it harder for small competitors. Why are you so dishonest?
Then why doesn’t any of the research agree with you? Why did some of the top researchers just release a paper saying that it’s actually more damaging to make these claims because it makes it more difficult to treat real addictions?
It’s incredible to me how people like you insist that things must be true based on “gut” and when the science proves you wrong you just dismiss them as “not being realistic.”