2021 Section 230 Year-in-Review

December 2020 ended with Section 230’s existence in jeopardy. Shortly before he insurrected a violent takeover of the U.S. Capitol, lame-duck President Trump demanded Section 230’s repeal. Trump even vetoed raises for our soldiers because Congress didn’t repeal Section 230 in the same bill. To placate Trump, Sen. McConnell drafted a Section 230 repeal bill. With so many power players treating Section 230 as a political football (and a useful distraction from the ongoing coup attempt), it wasn’t clear Section 230 would survive into 2021. See my 2020 Section 230 year-in-review for the depressing recap.

Section 230 had a better year in 2021, but that’s not saying much. Section 230 is clearly in trouble. Countless reform bills were introduced into Congress, all of them terrible. States continue to legislate as if Section 230 (and the First Amendment) doesn’t exist, and states are also lining up behind legal challenges to undercut Section 230 expressly or indirectly. And in the courts…well, almost every item in this year’s roundup represents a doctrinal retrenchment. The standard Section 230 case remains a quick defense win, but those are overshadowed by dramatic rulings increasingly favoring plaintiffs.

* * *

Justice Thomas Blasts Section 230 (Again). The Supreme Court dismisses the Knight First Amendment v. Trump appeal as moot, and Justice Thomas issued a concurrence where he free-associated–once again, without briefing–about Section 230, the First Amendment, and various ways that regulators could censor the Internet. #MAGA state legislatures around the country have responded to his dog-whistles with laws such as the Texas and Florida social media censorship bills.

Wisconsin Courts Can’t Agree on Whether Armslist Qualifies for Section 230. Two Wisconsin federal district courts said that Section 230 didn’t protect Armslist from personal injury claims by people shot using guns purchased there. The rulings conflict with the Wisconsin Supreme Court’s ruling in Daniel v. Armslist, which held that Armslist did qualify for Section 230. In both cases, the courts ruled for Armslist on other grounds–making the Section 230 conflict unnecessary.

The Yearbook Cases. The “yearbook cases” involve services that offer access to yearbooks behind paywalls. Using doctrines like publicity rights, plaintiffs are attacking how these services advertise by featuring people in the yearbooks and the vending of yearbook access. Some courts have circumscribed Section 230’s applicability: first, by saying that Section 230 doesn’t apply to publicity rights claims, and second, by saying the services uploaded the yearbooks, so the yearbooks aren’t third-party content. See, e.g., Sessa v. Ancestry and Knapke v. Classmates.

(Also, in Hepp v. Facebook, the Third Circuit reinforced that Section 230 doesn’t apply to publicity rights claims outside the Ninth Circuit).

The Online Marketplaces Cases. Cases involving the sale of items via an online marketplace raise two interrelated questions: are the marketplaces “sellers,” and if so, does Section 230 protect them from liability nonetheless? Courts have deeply split on the first question. Compare Amazon v. McMillan and Loomis v. Amazon. As for the second question, Section 230 is unlikely to help sellers because liability is based on their first-party activity. Online marketplaces are likely doomed if the SHOP SAFE Act is jammed through Congress; but the “seller” question potentially dooms them either way.

Second Circuit Drama Over Section 230(c)(2)(A). The Second Circuit issued THREE opinions in Domen v. Vimeo, one of the many cases seeking to force Internet services to carry unwanted and odious content. In the first opinion, the panel fully embraced Section 230(c)(2)(A) as the reason Domen should lose. In the second opinion, the panel walked back some of its most aggressive pro-230(c)(2)(A) language. In the third opinion, the panel stripped out the Section 230(c)(2)(A) discussion entirely and ruled against Domen on the substantive elements. This public drama indicates some serious internal friction among Second Circuit judges.

While the first opinion would have represented a major Section 230(c)(2)(A) victory, the net outcome was arguably more favorable to Section 230. It reiterates that Domen loses regardless of Section 230’s applicability. (A reminder: less than half of the account termination/content removal cases I’ve tracked rely on Section 230).

Ninth Circuit Drama #1: Enigma v. Malwarebytes. In 2019, the Ninth Circuit issued a troubling opinion in Enigma v. Malwarebytes, creating a murky “anti-competitive animus” exception to Section 230(c)(2)(B) that hindered anti-threat software vendors’ efforts to protect their customers. On remand, the plaintiff lost on the substantive elements. This is at least the third time when the Ninth Circuit rejected Section 230 only to have the case fail on substantive grounds. (The other two are Roommates.com and Doe v. Internet Brands). Of course Section 230 should fail when it doesn’t apply, but it doesn’t help anyone to create new doctrinal holes in Section 230 for cases that are destined to fail.

Ninth Circuit Drama #2: Lemmon v. Snap. The Ninth Circuit pulled the same trick in Lemmon v. Snap. The Ninth Circuit permitted the plaintiffs to proceed on a “software design” theory despite Section 230 (but only if the claim is based on the software design and not any third-party content attributable to that design). However, virtually identical claims against Snapchat failed on substantive grounds in Georgia, so this ruling probably created another unnecessary doctrinal hole in Section 230.

Ninth Circuit Drama #3: Gonzalez v. Google. I hoped to avoid blogging this ruling. The opinions are 132 pages (plus 35 more reprinting Judge Katzmann’s erroneous anti-Section 230 dissent in Force v. Facebook) from a badly split three-judge appellate panel. This ruling seemed primed for en banc review, but the Ninth Circuit declined to take it and instead left the messy tangle of opinions in place. UGH.

This ruling rolls up three lawsuits against social media services for allegedly materially supporting terrorists (the Gonzalez, Clayborn, and Taamneh cases). Those cases have failed across the country, including in the Ninth Circuit (see Fields v. Twitter, which the lead opinion barely cites), for a variety of reasons, including a lack of proximate causation, a failure of the substantive elements, the First Amendment, and Section 230. Indeed, two of these cases fail here. However, the court revived the Taamneh v. Twitter case because the “FAC alleges that defendants provided services that were central to ISIS’s growth and expansion, and that this assistance was provided over many years.”

The ruling exposed Ninth Circuit judges’ deep hostility towards Section 230. Judge Berzon says that, but for stare decisis, she would hold that Section 230 doesn’t apply to algorithmic activities (“activities that promote or recommend content or connect content users to each other”) because they don’t count as publication (UGH–this is a widely held but obviously wrong view). In a truly cringe-worthy and pro-censorship opinion, Judge Gould apparently says that if Congress won’t repeal Section 230, the courts should unilaterally do so through judicial activism.

In the Gonzalez case, the panel said that Section 230 didn’t apply to claims that the social media services shared revenues with terrorists because the “allegations are not directed to the publication of third-party information. These allegations are premised on Google providing ISIS with material support by giving ISIS money.” UGH. If true, Section 230 never applies to UGC services that pay authors for their content (because some of that content inevitably will be illegal, and that’s only time Section 230 would be needed). This is one of the most dangerous ideas in the SAFE TECH Act and, if successful, would radically reshape the UGC ecosystem. Worse, the panel still rejected Gonzalez’s claims on substantive grounds–yet another doctrinal incursion into Section 230 that helped no one. UGH.

Courts Are Baffled by FOSTA. FOSTA’s perniciousness has been proven many times. It’s hurt many communities and apparently has helped no one. Not surprisingly in light of the way FOSTA was drafted, FOSTA is a mess in courts, including:

  • Plaintiffs are asserting Section 230 does not protect content that seems far removed from sex trafficking.
  • Plaintiffs are suing defendants who are distant from the actual criminals, such as “tertiary” claims against Salesforce and Mailchimp. They are also suing Craigslist, which exited the industry in 2010, because apparently there are no statutes of limitation.
  • Courts cannot agree on whether an 18 USC 1595 claim gets around Section 230 based on the lower scienter requirements in 1595, or if it must navigate the higher scienter required by 18 USC 1591. Congress unquestionably meant the higher scienter standard, but judges are bending the statute to accommodate the plaintiffs.
  • FOSTA very clearly did not remove Section 230 protection for state sex trafficking civil claims, but the Texas Supreme Court misread the statute on this point anyways. (Facebook lost that ruling though it endorsed FOSTA. So shortsighted).
  • Courts are misapplying the Roommates.com “material contribution” standard to lower the Section 230 immunity so that state claims can become tenable despite FOSTA’s clear intent to keep Section 230 in place for those. This distortion of the “material contribution” standard could fold back into standard Section 230 jurisprudence, which means that FOSTA has the potential to break Section 230 generally.

For more insights into FOSTA’s failure in court, try reading this 5,400 word blog post.

Section 230 Celebrates Its Quarter-Century Anniversary. Section 230’s Silver Anniversary came on the heels of the December 2020 repeal attempts and the January 6 coup attempt, so many people weren’t in much of a celebratory mood (but I tried).

Given its venerability, a common anti-230 trope is that Congress could not have contemplated the current Internet when it passed Section 230 in 1996, so Section 230 is being used today in ways that were never intended. (As just one example from the Gonzalez lead opinion: “There is no indication the drafters of § 230 imagined the level of sophistication algorithms have achieved…Section 230’s sweeping immunity is likely premised on an antiquated understanding of the extent to which it is possible to screen content posted by third parties. There is no question § 230(c)(1) shelters more activity than Congress envisioned it would.”). This trope triggers me every time I hear it because Sen. Wyden and former Rep. Chris Cox literally told us in 2020 what they thought of Section 230. Spoiler: 25 years later, it’s doing exactly what they contemplated.