Senator Sheldon Whitehouse Has Really Bad Ideas About Section 230

Over the summer I got a copy of the new book from Lee Bollinger and Geoffrey Stone, two formerly staunch 1st Amendment supporters who have apparently decided to go back on their earlier views, with a collection of essays by a variety of authors about “social media, freedom of speech, and the future of our democracy.” Much of the book is maddening, because there are many essays from very famous people who should know better, but seem more than willing to reject the 1st Amendment because people said bad stuff online.

There are some pieces that are interesting and thoughtful, but the vast majority of them are incredibly frustrating. I had kind of blocked the book out of my mind, not wanting to revisit the frustration, but last week, Senator Sheldon Whitehouse decided to post a thread on Twitter promoting his chapter in the book, and presenting some of his ideas from his chapter — which was one of the most frustrating, by far, in the book. Bollinger and Stone obviously had an agenda in what they chose to put in the book, but Whitehouse’s chapter, in particular, would have benefited from having literally anyone who understands how Section 230 and the 1st Amendment intersect do an editing pass to call out some of the bullshit claims in it.

Unfortunately, after Whitehouse posted his thread, a lot of people cheered on his tweet saying “we should repeal Section 230, as it does more harm than good.” I criticized that tweet, and some people argued that it was unfair, because I didn’t deal with the entire contents of the thread. So, now I’ll go one step further and breakdown Whitehouse’s chapter in the book, and just how confused and wrong it is.

We should also note that Whitehouse has a bit of a history of being massively, dangerously wrong about the internet. He’s long been an extreme copyright maximalist, and was one of the initial co-sponsors of the initial bill that eventually became the SOPA/PIPA package. He’s been cozy with the top copyright industry lobbyists. A few years later, during a Senate hearing, he started making up nonsense about how a Google search he did lead him to The Pirate Bay and how this “criminal activity” (searches to websites) had to be stopped. He’s also been terrible on encryption, where he made up a story of a kidnapped girl and the only way we could possibly find her was to break into her phone. He also once pushed a bill to make the terrible CFAA even worse, and when called out on how much damage it would do, he blamed the “pro-botnet, pro-foreign cyber criminal caucus.” And, more recently, he pushed an obviously unconstitutional bill that would have required search engines to block searches that lead to information about illegal drugs.

Are you sensing a pattern yet?

Anyway, let’s get to Whitehouse’s book chapter. The premise is that (1) social media is bad, and (2) it’s because of Section 230, so (3) we should get rid of Section 230, and (4) while it’s possible that there would be some negative impacts of that, he has a “narrower” version of Section 230 that can replace it. Basically all four prongs of the plan are bad, and confused about reality.

The article starts out with the expected “social media is bad because the people on it say bad things” line of reasoning, cherry-picking from the usual pile of “social media bad” tropes:

Social media platforms—companies that facilitate information sharing through virtual networks—have shielded themselves more than any other media from responsibility for destructive content they house and propagate.

First off… what? Again, multiple studies have shown that cable news spreading bogus reporting has been way more responsible for misinforming the public. Indeed, there’s a whole book detailing, with very thorough data, how the viral stories go viral because of cable news, and not the internet. But, Whitehouse isn’t concerned with being factual. He’s here to burn down social media with disinformation.

Second, his “evidence” to support this claim is a footnote… linking to a Newsweek article that doesn’t even make the claim Whitehouse pretends it does. I’m guessing he didn’t think anyone would bother to check his footnotes, but the article is just about an advertising boss, who is not exactly an unbiased player, saying some nonsense about whether tech companies are “platforms or publishers,” which is a nonsense distinction.

They claim that their algorithms simply promote whatever is selected by the collective wisdom of the public, and that they lack the resources or expertise to identify and remove unlawful or untruthful content. But the truth is they are not neutral or incapable observers. Social media companies spread disinformation, exacerbate preexisting biases, and disseminate unlawful content because of deliberate, profit-seeking choices. These platforms choose how to structure their services; what content to allow or disallow; what content to promote; what ads to sell, and to whom; and how they connect advertising to the content users consume or create.

Then what’s your excuse for pushing this disinformation, Senator? This whole paragraph is misleading disinfo. No one claims that the platforms are “neutral,” but it is accurate that when you have an open platform for billions of people to communicate, some of those users are going to push garbage. No one claims that they can’t deal with any of it. Every platform puts in place rules and enforcement to try to deal with as much of it as they can, but the impossible part is actually being able to deal with all of it. And it’s not just because of “profit-seeking choices,” though if it were that, then rather than go after Section 230, why not go after Wall Street and its demands for short-term profit maximization for public companies?

And, again, it’s people who spread disinformation. As noted above, cable news does it much more impactfully than social media does. But Whitehouse isn’t looking to regulate cable news, because he recognizes the 1st Amendment problems with that. Unfortunately, he ignores them when it comes to social media.

Also, many of the other complaints are exaggerated or misleading in their own way. For example, the claims about “exacerbating preexisting biases,” the evidence there… suggests otherwise. As we wrote about last year, the evidence suggests that the internet actually doesn’t push people into echo chambers, but rather the reverse. The only “truth” to the idea that it exacerbates preexisting biases is by introducing people to other views leading some people to react negatively, and leading them to dig into their preexisting biases. That’s not the fault of social media. That’s the fault of bad education and people who are scared of new and different ideas. Kinda like Senator Whitehouse who seems to keep exacerbating his preexisting biases against the usefulness of the internet.

So, we’re already off to a not great start, but it quickly gets much, much worse. Because, while lots of people insisted that Whitehouse had a serious plan to “repeal and replace” Section 230 with “something better,” in the chapter, he suggests a total repeal would be the best overall solution:

I should say at the beginning that I would support simply repealing Section 230 and letting courts sort it out. This has the advantage of legislative simplicity and speed. It also minimizes the hand of Congress in an area that relates to speech, where our own political motives—whether of incumbency or party—create their own hazards. Better to minimize Congress’s hand in this.

So, yeah, I agree with minimizing Congress’s hand in this, and he’s correct about the problems of Congress meddling in areas relating to speech, where political motives may take control, but it’s utter madness to suggest that repealing it and “letting courts sort it out” is any kind of reasonable solution.

First, to argue that it’s faster is similarly madness. If we simply removed Section 230 today, tomorrow the courts would be flooded with tons of ridiculous and frivolous cases, blaming social media for basically everything, and it would take years to sort out the mess, and it would be incredibly costly for everyone sued. And while totally out-of-touch Senators like Senator Whitehouse can dismiss this concern by saying that the big tech companies have plenty of money and lawyers to deal with this, it ignores the fact that Section 230 protects tons of smaller sites, that would also get caught up in this, and many would go out of business trying to defend these frivolous lawsuits.

It is the height of elitist nonsense to say “let’s just create massive turmoil” for every website on the internet, because I don’t like the internet. But that’s what Senator Whitehouse does here. It’s obnoxious. It’s based on fallacious assumptions, and it’s dangerous.

Senator Whitehouse suggests that this mess would be relatively short-lived:

Most of the questions that would come up in court post-repeal would find ready answers in existing legal doctrines, with familiar structures and duties. Repeal is not a ticket to an alien legal environment; it’s actually a return to established legal norms.

Tell me you’ve never been a small business sued by frivolous grifters without telling me you’ve never been a small business sued by frivolous grifters. As far as I can tell, Whitehouse, whose great great grandfather was a railroad magnate, and whose parents and grandparents were career diplomats, has never worked in the private sector at all. He got a law degree, became a law clerk, and then went straight to work in the government. It is the height of hubris to suggest that a ton of small, struggling businesses should have to go through years of expensive litigation because you’re mad that some people on the internet aren’t nice.

From there, Whitehouse insists that without Section 230, “disinformation” directed at individuals would be solved by lawsuits:

Where disinformation targets an injured individual, liability law will usually clean up the mess.

And here we begin to realize the root of the problem: Senator Sheldon Whitehouse does not understand the 1st Amendment. Most disinformation is protected speech under the 1st Amendment, Senator. There are some exceptions, but they are pretty limited, because we protect freedom of speech. What’s lacking from Whitehouse’s analysis (and this becomes a bigger deal later in the piece) is the lack of understanding that there is no underlying cause of action for most disinformation, because it’s protected under the 1st Amendment.

He goes on a weird, and somewhat misleading, history of Section 230, followed by a very, very, very misleading explanation claiming that social media sites have no interest in stopping the spread of disinformation on their platforms. This is simply untrue, and fails to reckon with the very real challenges and tradeoffs in trust and safety, and how these companies have tried to balance those different trade-offs.

Like most interlopers with zero experience in the field, Senator Whitehouse writes as if there’s some easy solution. Much of this part of his chapter reads as a Senator with no experience in healthcare saying “the solution is easy, we just need to ban health insurance, and everything will sort itself out.” That’s not how anything works, and your “solution” completely ignores all the work of tons of experts who have spent decades actually making the internet better.

And, again and again, Whitehouse seems to struggle with understanding what is protected speech under the 1st Amendment. In a section of his chapter about “illegal content,” he repeatedly presents examples of content that, while bad, is not illegal. He talks about disinformation and misinformation. He talks about anti-vax content. He talks about false COVID info. He talks about “climate denial.” We can all agree that this information is problematic, but it’s all very much protected.

And… removing Section 230 would actually make it that much more difficult to deal with. Section 230 is what allows different companies to experiment with different approaches and figure out what works. It’s why all of the major social media companies have implemented increasingly beneficial policies for handling this stuff, and that they can adapt rapidly. Without it, every single change would require a review by the legal team, who is risk averse, and would greatly limit the ability of companies to keep ahead of malicious actors on their sites.

This is obvious to anyone who has worked on these issues, which does not include Senator Whitehouse.

Whitehouse continues to insist that companies aren’t even trying to do anything about these issues, which is an insult to their trust and safety teams who actually do work incredibly hard to put in place and effectively implement policies and enforcement to improve the various websites. Whitehouse’s entire chapter is an insult to all the work those teams put in.

Finally, towards the end of the chapter, after he suggests a full on repeal, and after making a ton of false claims (disinformation?), Whitehouse does admit that just getting rid of Section 230 might have some negatives. It’s almost like halfway through the chapter, someone else took over writing it. He notes that repealing 230 would leave a lot of uncertainty. He notes that misinformation often isn’t legally actionable (something he then forgets later in the chapter). He even admits that Congress can’t just outlaw misinformation because of the 1st Amendment (though he sort of hides the ball by not directly mentioning the 1st Amendment, and instead just talking about “strict scrutiny”):

Congress can’t readily solve these problems by creating new causes of action. Causes of action based on the content of speech — for example, a new cause of cation for knowingly publishing misinformation online — will be subject to strict scrutiny in court. Many statutes seeking to criminalize cyberbullying or other online speech have been struck down on vagueness grounds.

Vagueness isn’t the problem here. The 1st Amendment is.

He also points out, correctly, that “unlimited liability could privilege wealthy special interests.” That, of course, is the point that some of us keep trying to highlight. If you want to lock in Facebook and Google, get rid of Section 230. They have the lawyers and the bank accounts to deal with the fallout. No one else does. Whitehouse actually is correct in saying this… but then seems to immediately forget about it and not care about it:

If Section 230 is repealed without additional guidelines, an already unbalanced information ecosystem could be unbalanced further as platforms yield to legal pressure from big, deliberate manipulators of information. Powerful special interests can bring lawsuits they are unlikely towin in order to scare off social media companies in terms of how they police certain content.

[…]

The ultimate success of a lawsuit, however, may not matter to well-funded interests with the means to threaten nuisance suits, and there are other doctrines of tort law that could be used to frame a dispute or a threatened dispute.

And, of course, the benefit to the biggest companies is mentioned as well:

Trillion-dollar social media companies could be beneficiaries as well as victims of nuisance litigation. With an abundance of resources at their disposal, Google and Facebook can easily afford to litigate. This gives them an incumbency advantage: New social media startups cannot afford to spend millions of dollars on litigation. Startups also can’t afford to spend millions of dollars developing automoderation mechanisms.

This is… all accurate. Yet most of the chapter, both before and after this part, completely ignores this. It’s almost as if he handed off his pen to someone more well informed, and they added this section which is then otherwise ignored.

You’d think that whoever wrote the parts above would then recognize why Section 230 is actually useful. But instead, it leads Whitehouse to finally laying out his “proposal.” Which is basically… a DMCA for misinformation, plus some additional transparency mandates.

The best solution would be for Congress to require a “notice-and-takedown” systems removing Section 230 protections when a company willfully refuses to remove unlawful content. As part of this system, major social media platforms should maintain an “acceptable use” policy, explain how the platform enforces its content moderation policies, and describe the methods of reporting content or speech that violates policies or other laws. They should notify users when their content is taken down, and give users a forum for appeal if they they’ve been wrongly removed or if the company has failed to act.

All of these ideas have been suggested elsewhere, and all of them have very, very serious challenges and tradeoffs Whitehouse does not grapple with. First, we already have a DMCA notice-and-takedown provision in copyright that is massively abused to try to silence speech. A big study of how well notice-and-takedown works in everyday practice… shows that it doesn’t. As that paper notes, they found “surprisingly high percentages of notices of questionable validity.” Expanding the notice-and-takedown system much more broadly without recognizing how the existing one already puts a burden on protected speech should be a non-starter.

Also, note the sleight-of-word trickery in this paragraph. He notes that the notice-and-takedown provision should cover both “speech that violates policies or other laws.” But… policies are not laws. There’s an important difference there. Speech that violates laws is an issue, and almost every website will actually remove it once it’s proven that the speech is violative. The problem is that people like Whitehouse assume that there’s some obvious marker of “illegal speech” as compared to speech that is legal, but there is not. There’s a reason why we have a court system that takes a lot of time and back and forth with hearings and trials and evidence and juries to determine if something breaks the law. This proposal ignores all that, and assumes that a website can just tell, based on a report, if some speech breaks a law. That’s not how any of this actually works.

And if we’re just talking about internal policies, well, companies already try to remove content that violates those policies once they’re informed of it, but the problem is that people with zero experience in this field (like Senator Whitehouse) don’t understand that it’s rarely as obvious as they think whether or not content actually violates policies. Sometimes it is, but the biggest issues and the biggest challenges are that so much content is in a gray area where it’s not really clear if they break the rules or not, and many judgment calls are made, often by understaffed, overworked teams with little time to judge the context or nuance of the issues at play.

Whitehouse doesn’t care. He assumes, incorrectly, that companies are deliberately ignoring this content, when the reality is that they’re trying to figure out ways to enforce policy across billions of people, in which every scenario is impossible to fully understand, and millions of judgment calls are made every day.

As for the demands for an appeals process, and clear explanations, again, that is massively burdensome for smaller players. Techdirt removes between 500 and 1000 spam comments per day. I should need to contact each spammer to let them know, and share with them our “appeals” process? Fuck that.

He has a few more suggestions that are just as confused, and just as untethered from reality. For example, he insists that Section 230 should not apply to algorithmic recommendations (which the Supreme Court might solve for him this term). But that basically would destroy search on any website, let alone many other important, and useful algorithms.

He keeps forgetting that most misinformation is constitutionally protected. For example, he says:

The threat of legal liability, for example, could make Facebook and other companies more likely to adopt measure that stop the spread of misinformation even if they also reduce user engagement, as failing to act would carry its own financial risks.

But, as even he admitted earlier in the chapter, misinformation is mostly protected under the 1st Amendment. So what possible cause of action could there be for failing to remove misinformation? Also, this sentence completely ignores the other statements in this very chapter highlighting how powerful interests will use the threat of litigation to pressure companies into hiding legitimate information they want hidden, as well as the vastly different scenarios for smaller competitors that don’t have Facebook’s legal team and bank account.

The whole chapter is weirdly disconnected from reality. There is that weird bit in the middle that accurately highlights many (but not all) of the problems with his own proposal, which he then completely ignores.

Senator Whitehouse, who has never worked for a private company and has no experience with the internet other than proposing and supporting ridiculously bad regulations for the internet, does not understand the problem at all (he misrepresents why content moderation is such a challenge, blaming it on profit-seeking, rather than the complexities of human beings and society). And his proposed solutions have so many negatives that he fails to grapple with. And, even when he does grapple with some of the downsides to his proposals… he then just ignores them as if he hadn’t even mentioned them.

It’s the worst of political nonsense. It’s grandstanding on an issue he doesn’t understand, with a solution that will not work and will make the actual issues that much worse. This is not good policymaking at all.