Fifth Circuit Once Again Disregards Supreme Court Precedent and Mangles Section 230–Free Speech Coalition v. Paxton

Texas passed a law (HB 1181) requiring pornographic websites to age-authenticate all users and then prevent minors from accessing online porn. If this sounds familiar, it’s because Congress passed functionally identical laws twice: the CDA in 1996 and the COPA in 1998. Both laws were struck down as unconstitutional. But to the Texas legislature, directly-on-point adverse Supreme Court precedent are just an invitation to further negotiations. Why not go ahead and pass an equivalent law? As my grandma might have said, if you don’t ask the Supreme Court to overturn their long-standing prevailing precedent, the answer is always no.

To help Texas along with its Conlaw resistance efforts, the Fifth Circuit–a known graveyard of the rule of law–cherrypicks Supreme Court precedent to embrace or disregard as it sees fit. That’s how the Fifth Circuit reaches the risible conclusion that HB 1181 only triggers rational basis review. There’s no point in blogging that conclusion as if it’s legitimate judging.

(If you care for the details, the panel majority claims the 56-year-old Ginsberg opinion, which dealt with offline retailers, governs the Conlaw analysis of the Texas law instead of the squarely on-point 1997 Reno v. ACLU and 2004 Ashcroft v. ACLU opinions, both of which dealt with the Internet. Judge Higginbotham, dissenting in this case, makes appropriate rebuttals).

This post instead focuses the court’s treatment of Section 230, not because the Fifth Circuit got it right (it didn’t), but because this is another entry in the Fifth Circuit’s increasingly unstable Section 230 jurisprudence.

The court starts out with a misreading that Section 230 only applies to removal decisions, not leave-up decisions:

Texas correctly suggests that Congress enacted § 230 to shield against liability for removal, but not promulgation, of ‘offensive material.’…But plaintiffs seek to…make (c)(1) a shield for purposefully putting “offensive material” onto the Internet.

(Note the slippery word “promulgation”–can you think of any more accurate synonyms the panel majority might have used here?)

Leave-up/removal decisions are the exact same decisions. Imposing liability for leave-up decisions necessarily imposes liability for removal decisions. Also, Texas has tried to functionally eliminate services’ discretion to make removal decisions via its social media censorship law–and the Fifth Circuit upheld that law too. Putting the two together, the Fifth Circuit indicates that 230 only applies to removals, but services can’t actually make any removals, so…..I guess Section 230 doesn’t exist in Texas?

The court’s reference to “purposefully” also implies that this panel majority supports a scienter-based exception to Section 230, i.e., services qualify for Section 230 if they lack scienter, but show defense scienter and 230 is toast? I discuss how Section 230 eliminates scienter-based inquiries, and why that’s a good thing, in this paper.

Unfortunately for the panel majority, the pesky 2008 Fifth Circuit ruling in Doe v. MySpace read Section 230 broadly. The Fifth Circuit recently came within 1 vote of repealing that case en banc, and I wrote at the time: “any future Section 230 case heading to the Fifth Circuit faces an extreme risk of judicial activism to overturn the existing Fifth Circuit precedent and disrupt decades of Section 230 jurisprudence.” Indeed, this panel majority has already disregarded SCOTUS precedent, so it’s not going to feel constrained by Fifth Circuit precedent either.

Having previously said that Section 230 only protected removal decisions and not leave-up decisions, the panel majority whipsaws the reader and says Section 230 actually does protect leave-up decisions:

Like cellphone service providers, interactive computer service providers cannot be held liable for harmful communications that they fail to remove. But liability under H.B. 1181 is not like liability under a negligence claim. It is not reliant on the harm done by third-party content. It imposes liability purely based on whether plaintiffs comply with the statute, independently of whether the third-party speech that plaintiffs host harms anybody. As Texas puts it “if a minor circumvents age verification, ignores the health warnings, and is subsequently harmed by third-party content or resulting offline conduct,” Section 230 would kick in and bar liability. That is the nature of Section 230’s protections: to protect a provider from speaker-liability stemming from the speech it hosts.

Among other defects, the analogy to “cellphone service providers” is trash. Cellphone service providers are not legally permitted to listen into their subscribers’ conversations, so they have no power to “remove content” at all. Also, cellphone service providers don’t host speech.

As I said, don’t try to overthink this. The panel majority is playing Calvinball. At one point, the panel majority declares without a whiff of irony that “Plaintiffs mislead the reader.”

The panel majority never says it so clearly, but I think it’s trying to say that the age authentication mandate only regulates the services’ conduct, and thus it doesn’t impose liability for third-party content. A conduct/content distinction is often illusory in the publisher context because editorial decisions pervade the “conduct” of publishing. Furthermore, as I read it, HB 1181 applies equally (as did CDA and COPA) to a website’s publication of its own content and third-party content. The statutory drafting tries to mask that fact by putting the age authentication obligation front-and-center (which itself ought to be enough to trigger First Amendment scrutiny). However, fundamentally, the statute imposes liability for services for publishing third-party content to underage viewers, and Section 230 clearly should apply to that aspect.

This opinion will be appealed to the Supreme Court, alongside other cases over statutes imposing mandatory age authentication. The pro-censorship forces have been angling for an opportunity to challenge Reno v. ACLU and the COPA caselaw, hoping that the Supreme Court will forget or overturn that precedent. This highlights the stakes of this and the other cases on their way to SCOTUS: do the foundational Constitutional law principles that have fostered the Internet’s success over the past 25+ years still apply, or have the rules since changed and opened the door to rampant government censorship? The Internet’s fate–and perhaps the fate of free speech in our country–hangs in the balance.

Case Citation: Free Speech Coalition, Inc. v . Paxton, 2024 WL 982225 (5th Cir. March 7, 2024). Prior blog post.