Speech

Amy Coney Barrett is bringing up the Hotel California clause yet again, and I’m actually grateful — she points out the plain language of the law seems to specifically say you can’t ban Texan users. Nielson says it’s a conditional rule. “If you choose to do business in Texas, then this provision kicks in,” but “if you don’t want to do business in Texas at all” you’re okay to geofence the state. “You can’t darn well discriminate” against Texas users if you operate in there, he says.

Or in his words, “intensely competitive” — which he says doesn’t change the fact they could be considered common carriers, so the existence of multiple social networks shouldn’t save those from regulation either. But there are basically three mobile networks in the country! It’s not an outright national monopoly, but it’s a pretty consolidated space.

Nielson says it’s not an accurate read of the law, but Roberts expresses apprehension about whether it would really be possible to pull out of the state in a way that satisfies its requirements. “I don’t see how they can wall off Texas,” he says.

Amy Coney Barrett suggests HB 20’s scope is more limited than Florida’s equivalent law, saying it only covers the “classic social media sites” like Facebook, not platforms like Etsy. Nielson agrees with her, and weirdly nobody brings up Wikipedia — whose operators have expressed concern they’d be covered by the law.

Nielson is defending HB 20 now. A group of legal scholars, he points out, worry that striking down the Texas law could make tech company regulation in general impossible — although even they call the Texas law “dangerous.”

Clement mentioned what he calls a “Hotel California” provision of HB 20, which he interprets as a ban on companies pulling out of Texas if they can’t meet its legal burdens. Brown Jackson questions whether that’s a reasonable interpretation of the law; she seems less convinced it’s the right read.

A lot of today’s fight has been about metaphors, and Alito is questioning whether some of them make sense — a newspaper in NetChoice’s view, and a common carrier like a telegram company in the states.’ Clement points out that the court has regulated the internet specifically before in cases like Reno v. ACLU, which struck down most of the Communications Decency Act. Unfortunately, that doesn’t really clear up the metaphor question.

Clement fields a question from the court on why Texas’ requirement to explain social media takedowns is more problematic than EU laws requiring some level of consistent moderation and explanation — he argues that Texas’ individual response provision would be “incredibly burdensome.”

Several justices (both liberal and conservative) have seemed sympathetic to the idea that private companies can engage in harmful censorship, with Alito referring to the possibility as “Orwellian.” But Kavanaugh keeps pushing back on the premise. “We don’t want the state interfering” with these private entities, he says, even if they’re powerful.

Speaking to US Solicitor General Elizabeth Prelogar, he suggests Section 230 is conditioned on the idea that web platforms are common carriers. That’s a weird take on Section 230, which is regularly applied to small blogs and online newsletters — and was passed partly to avoid punishing services that moderated content to be “family friendly” rather than acting as neutral conduits. Prelogar, for what it’s worth, disagrees with Gorsuch.

Brown Jackson brings up a common critique of NetChoice’s position: the internet has become the public square, and large social media companies play a huge gatekeeper role, so why not regulate them? NetChoice’s Clement argues that web platforms rise (he mentions TikTok) and fall (sorry, X) quickly enough that it’s clear there are lots of avenues for speech online.

It’s been brought up by other justices in passing, but Thomas really keeps harping on whether striking down this law should mean Section 230 is unjustifiable. He’s bringing up some questions that are interesting but also pretty tangential to the core case, like whether “deep learning algorithms” are platform speech. (Clement says... probably.) “What is the algorithm saying?” Thomas asks.

Amy Coney Barrett expresses discomfort with whether striking down the Florida law would prevent any regulation of web platforms as common carriers. Clement’s reaction is effectively that the law is so bad that it should simply be struck immediately and any future questions sorted later, but other justices break in and question that logic.

So why can’t a law focus on specifically regulating some of the world’s biggest communications services? Clement vociferously disagrees, citing Minneapolis Star Tribune Co. v. Commissioner among other cases. We’re getting into the most complicated question of the case here: how far should the Supreme Court go in protecting “Big Tech” companies from regulation overall, not just striking down these particular laws?

After reminding us he’s older than the internet, Thomas makes the claim that sites have been claiming Section 230 protections as neutral “conduits,” accusing them of hypocrisy for saying they hold editorial standards now. NetChoice’s Clement disagrees — saying they’re simply not supposed to be treated as the publishers of specific user-generated content.

Etsy keeps coming up as an example of a smaller user-generated content platform that could get hit by Florida’s law, and Whitaker keeps suggesting that its moderation is fundamentally non-speech-related and therefore irrelevant to the debate. As Platformer lays out, that’s simply not true — Etsy’s moderation of speech around Israel and Palestine has in fact been incredibly fraught.

Coney Barrett asks whether automated decisions are fundamentally different from conventional editorial judgments. Whitaker tries to bring up Twitter v. Taamneh, in which sites said their automation in a particular case (involving terrorism) didn’t present a viewpoint.

Brett Kavanaugh notes Whitaker’s opening statement doesn’t mention that the First Amendment is conventionally focused on government suppression of speech, not private speech decisions. Whitaker says there’s still a larger First Amendment interest in protecting freedom of speech from censorship by other parties.

Clarence Thomas complains that there’s a lack of specificity in the discussion about what’s covered, and Samuel Alito pushes on whether the law could regulate “expressive” conduct that should deserve First Amendment protection.

Brett Kavanaugh asks Florida attorney Whitaker about one ongoing question from critics: would saying the First Amendment doesn’t prevent forced speech on websites undercut the rights of all kinds of other businesses that deal with speech?

“But the one thing I know about the internet is that its variety is infinite,” Sotomayor continues. “So at what point at a challenge like this one does the law become so generalized, so broad, so unspecific really that you bear the burden of coming in and telling us exactly what the sweep is?”

Texas and Florida’s anti-moderation laws are explicitly pro-Republican, but as my colleague Lauren laid out last week, the lawsuits over them could have bigger tech regulation fallout. While we’re waiting for oral arguments to start, check out The Atlantic and The New Republic for a couple of pieces on the states’ counterintuitive bipartisan appeal, plus some countertakes from Lawfare and Techdirt too.

The Washington Post has been pushing Meta on what its recent decision to stop recommending “political” content and “social topics” on Threads and Instagram means, and it’s culminated in this illuminating quote from Meta spokeswoman Claire Lerner:

Six months after kicking off an initial Mastodon trial that saw it launch its own instance on the federated platform, the UK’s public broadcaster is not just extending the experiment by another six months, it’s also “planning to start some technical work into investigating ways to publish BBC content more widely using ActivityPub.”

Child sexual abuse material is a well-known exception to the First Amendment, but the law around AI-generated simulations of it is vastly more complicated. Lawfare’s new analysis addresses some of my longstanding questions — along with others I hadn’t even thought to ask, including the status of tools that accidentally train on CSAM. It’s long, but if you’re interested in how AI will test the criminal justice and legal systems, absolutely worth the read.

Oral arguments in Moody v. NetChoice and NetChoice v. Paxton — respectively concerning whether Florida and Texas can take over social network moderation — were finally put on the docket. Will we see a continuation of last year’s surprisingly circumspect internet law discourse? Or will Clarence Thomas decide the time is right for the crackdown he so craves? We may find out in a couple months!

Legal blogger Ken White (aka Popehat) isn’t outright ditching Substack over its decision to keep monetizing Nazi content. But he finds its claim of being a principled upholder of free expression, rather than a company pulling a widely recognized branding trick, a bit risible:

It’s from New York, and it would require large social media companies to let third-party services do things including blocking or muting content on behalf of users — basically mandating a certain level of free, open API access. Mike Masnick has some thoughtful analysis:

As 404 Media notes, an investigation into alleged “hacker-for-hire shop” Appin has been removed — Reuters says temporarily — after what Reuters calls a preliminary court order that the news outlet is fighting. There’s no clear evidence of factual errors here, which might make this a case of de facto press censorship in India. You can still read the feature here.