The State Lawsuits Against Internet Companies Will Continue Until Humans Improve
The big Meta settlement with a bunch of states got most of the headlines, but state AGs have gone all in on suing tech companies of late, and it shows no signs of slowing down. Last week things went ballistic when the judge in New Mexico's case (New Mexico being one of the few states that refused to join the others in settling) ordered Meta to pay a $25 billion penalty. As Daphne Keller points out, a rough back-of-the-envelope calculation shows that $15 billion of that judgment covers the same claims that all the other states together settled for just $459 million (though Meta has said it will fight the ruling).
As we discussed earlier, TikTok also settled with Alabama in a separate case. But the cases just keep coming. Rhode Island just sued Meta (again), this time claiming it allowed scam ads to run rampant on the site. And a group of state AGs (mostly using the same outside law firm) sued OpenAI separately (same basic lawsuits filed individually in each court). TikTok, Snap, YouTube, all face additional similar lawsuits as well.
Many people are cheering these lawsuits on, and you can understand why. These are some of the most hated companies on the planet right now. No one trusts the big Silicon Valley companies and their broligarch CEOs who seemed willing and eager to fall in line (literally) behind the fascism of Donald Trump. Not to mention the enshittification. Oh so much enshittification. And then add the relentless attempts to force AI tools into places people clearly don't want them.
People were understandably and reasonably fed up. And they wanted blood. Massive verdicts deliver it.
But as I've warned, we shouldn't be too quick to celebrate these kinds of things. Indeed, we may quickly be learning exactly why I was so concerned with those early verdicts against Meta earlier this year. As I noted at the time, some of those cases were the kinds of lawsuits that Section 230 was supposed to protect against - and for good reason. Other cases (like the latest New Mexico ruling) do involve statements by the company (thus they're not Section 230 issues) but are still equally problematic, as we'll discuss below.
First, most of the product liability, negligent design, and kid safety cases really shouldn't even exist under Section 230. Effectively all of them are predicated on the fact that some people do bad things, some (not all) of which were loosely connected to the internet services that are being sued, under the theory that the internet companies should have prevented the bad things from ever happening. But the entire premise of Section 230 was that if you hold internet platforms liable for every bad thing that happens on them, they will massively crack down on speech online, to avoid facing liability. It turns the internet from being a global communications network... into a bland broadcast" tool, where only approved content gets shared.
It also makes it almost impossible to run an online platform that allows third party speech unless you have massive resources and a building full of lawyers. In other words, it means that only the largest internet companies can survive, while making it way harder for there to be any new entrants in the market.
There's a reason that State Attorneys General spent many, many years pushing Congress to amend Section 230 to carve out the enforcement of state laws. They never actually got that. But they may have gotten something better. A bunch of judges who decided that state AGs and the courts can simply ignore Section 230 and hold internet companies liable for things their users do, so long as you claim the bad things were due to product design" or negligence." The Ninth Circuit recently eviscerated the key part of the law that was supposed to stop exactly these kinds of cases.
It's the get around 230" trick that collapsed the entire law.
As for the New Mexico case (and some of the new Rhode Island case), they are, in theory, predicated on statements by the company. But when I look at the New Mexico ruling (even setting aside the size of the judgment), I can't figure out how it makes any sense. The jury verdict is based on the claim that public statements by Meta and some of its executives were willfully deceptive." And, as someone who doesn't trust Zuckerberg or Meta at all - and who has called out some of his willfully deceptive statements in the past - I was primed to side with the jury.
But look at the statements" in question. I'm having trouble seeing how any of them are actually deceptive in any meaningful way. I'm not going to post them all, but they all seem mostly true, even if the public believes otherwise. Take for example:
We don't sell your data and we don't tell advertisers who you are. What we do is allow advertisers to reach people interested in their products."
This statement (by Sheryl Sandberg in this case) is... accurate. I know that, broadly, people often claim that big tech sells your data" but that's never actually been true. Their business model has always been based on selling advertisers access to your feed based on your data. That's a very different thing. What the company does is allow advertisers to target you based on the data Meta has about you, but the advertiser never gets or sees any identifying information from you or data about you. There are data brokers out there that do sell that kind of data (and those should be wiped off the earth, as far as I'm concerned), but that's never been any of the big tech companies' business. Indeed, one of the statements the jury found deceptive was Mark Zuckerberg accurately describing this dynamic:
There are other internet companies or data brokers or folks that might try to track and sell data, but we don't buy and sell."
That's an accurate statement! But the jury found that this one statement alone generated 1,386,648 violations of the law! What?!? How that and all the other violations" (almost 44 million of them!) came about is also highly questionable. It assumes that each New Mexico resident who might have come across one of the statements counts as an actual violation under the law:
It assigned 2.1 million violations to each of 11 statements that appeared in major news outlets and 1,386,648 to each of 15 statements made through company channels, including its newsroom, executives' Facebook posts, earnings calls and a developer conference, for a total of 43,899,720.
According to reports, 2.1 million was the population of New Mexico in 2020. 1.4 million was the number of New Mexico residents with active Facebook accounts at the time. So basically the jury claimed that each person who could have seen one of those statements because they appeared in a major news outlet... equals a violation of the law. And for the statements made on Facebook/Meta properties themselves, they assume every person with an account could have seen it, making it another violation. That calculation seems... bizarre. Lawyers are already saying they doubt it will hold up on appeal.
Even worse, Meta execs also got dinged for general statements about how they try to take down misinformation and don't profit from misinformation:
Well, if we see harmful misinformation on the platform, then we take it down. It's against our policy. So the 18 million number that I shared is the number of pieces of content that we've seen on the platform that we take down. Now, do we catch everything? Of course there are mistakes that we make or areas where we need to improve, but that's the best number that we have in terms of what we've seen and what our systems have been able to detect."
That's Zuck again, and that statement is also... pretty accurate? They certainly do miss some harmful content, but that is the nature of content moderation, especially when you have billions of users around the globe. I don't see how that claim can be deceptive.
In court, New Mexico apparently argued that this was deceptive because of Meta's Xcheck program. But as we were just discussing last week, there was a perfectly valid reason for Xcheck. It was the program that was put in place to deal with high profile accounts. Those were the accounts where Meta recognized that if they made a mistake, and took down those accounts incorrectly, it would make the company look really bad. So they created a system where no action would be taken on those accounts without a secondary human review. Given the situation, this sounds reasonable. It's what most companies would do facing similar challenges (and, it is, in fact, what most companies have done).
But when Frances Haugen released the details of this program, it was reported on (misleadingly) as Meta allowing the elite to violate its rules." But that's not what the program was for. It was to make sure there was additional human review, even if that review sometimes took too long.
But New Mexico presented it to the jury as Meta claiming to take down harmful misinformation while secretly using Xcheck to let big accounts spread it, and the jury (and now the judge) punished Meta accordingly.
I'm honestly not sure how any internet company avoids facing a similar risk down the road. When an AG can take broadly true statements, strip them of context, and call them deceptive (when the only deceptive thing is the AG's presentation), how can any company survive such lawsuits?
And, with the chance to get a $25 billion award from it, these lawsuits are going to be everywhere. It could even be worse, given that New Mexico actually asked for even more, and the judge said he could have ordered a penalty over $200 billion. And, again, these are for basically true statements! How do you run an internet company with that kind of threat of liability over your head?
Even if you look at the strongest elements of New Mexico's case against Meta, I'm left a bit baffled. The key part of the case resolved around the claim that You have control over how your information is shared," which was a statement from back in 2010. The court in New Mexico claimed that was deceptive... because of the FTC's $5 billion fine of Meta, over the Cambridge Analytica scandal. That fine hinged on the same claim, with the FTC arguing that it was misleading by Meta to claim that because anyone could also share their friend graph, and since that graph includes information about other people (namely that you are connected to others), they argued that revealed private info.
But, as we noted at the time, that doesn't seem like that big a deal? Knowing who is in your friend graph from a social media site doesn't really seem like violating privacy. Friend graphs are... just how social media works. There's also no evidence that anyone was directly harmed" from the ability to access a friend graph this way.
Finally: the FTC fined Facebook for this already, and did so at $5 billion for the entire country. Why does anyone think that New Mexico alone deserves 5X that amount?
Combined with the judgments and the settlements, these cases are going to keep piling up, and they're going to make the internet worse in almost every way. If you thought enshittification was bad before, wait until politicians and judges are the true product managers for the internet, driven by headline seeking and profit maximizing Attorneys General who will view suing tech companies as a way to fill state coffers.
People can celebrate the fact that big internet companies we all hate are losing cases and having to pay, but the end result won't be a better internet. It will be one driven by politicians looking to shake down tech companies any time anything bad" happens online. That's going to lead to way more restrictions on the internet. Way more identity verification to do anything. Way less innovation (and new offerings will have to go through an exhaustive legal review - not to make them better, but to make them less likely to end up in court). It's also going to lead to way less transparency. The New Mexico verdict means that even accurately describing how your site works can lead to a massive verdict against you, so companies are going to tell execs to shut up completely. This will also mean very few new competitors - except those with massive bank accounts. The broligarchs will be able to afford the liability risks of new services. Everyone else will be priced out.
This was the world that Ron Wyden and Chris Cox sought to avoid when they authored Section 230 30 years ago. But it's here now - even without the repeal of Section 230 - and we're going to see just how bad that is for the internet.