Ninth Circuit Panel Goes Out of Its Way to Question Section 230–Doe v. Meta
[I’ve been sitting on this ruling for almost a month because blog posts like this are time-consuming and emotionally draining to write. It may not look it, but this post took about 6 hours to write.]
This case involves a terrible tragedy: genocidal violent attacks on the Rohingya minority in Myanmar at the beginning of the 2010s. The plaintiffs sued Facebook for its role in the attacks, such as its alleged algorithmic turbocharging of rage content posted by its users.
Despite the tragedy, the facts sound like a straightforward Section 230 situation. However, the lower court didn’t rely on Section 230 to dismiss the complaint. Indeed, Section 230 isn’t mentioned a single time in the district court dismissal, part of why I never blogged that opinion. Instead, the district court dismissed the case solely on statute of limitations grounds (“the Court determines plaintiffs’ claims, having been brought in 2021, were filed outside the applicable two-year statute of limitations”) without mentioning Section 230 even once. The panel acknowledges that this case did not present itself to the Ninth Circuit as a Section 230 case: “Because the district court dismissed Plaintiffs’ claims as untimely, it did not reach the Section 230 issue.”
Given that Section 230 is nowhere to be found in the lower court opinion, the Ninth Circuit could have affirmed the lower court on statute of limitations grounds. Or, if it disagreed with that ruling, it could have reversed the lower court’s ruling and remanded the case to the lower court to evaluate other aspects of the case, such as Section 230.
This panel did neither. The Ninth Circuit panel’s opinion doesn’t engage with the statute of limitations issue at all, i.e., it doesn’t indicate if the lower court was right or wrong on that topic. Instead, the Ninth Circuit panel requested the parties file supplemental briefings on Section 230 grounds (remember, the plaintiffs couldn’t initially appeal on Section 230 grounds because the lower court didn’t mention Section 230 at all) and then conducted its own de novo application of Section 230 without any guidance at all from the lower court.
In other words, the Ninth Circuit panel didn’t have to discuss Section 230. IT WENT OUT OF ITS WAY TO DO SO. The panel then raised Section 230 on its own initiative and then criticized Section 230’s application as a problem–even though Section 230 didn’t dictate the outcome at the lower court. The panel’s issue-seeking is a blazing red flag of judicial activism.
(Also, the panel opinion isn’t very transparent about why it chose to discuss only Section 230 and entirely ignore the lower court’s statute of limitations ruling. It treats Section 230 as a critical-path item without disclosing that the panel went off-road to address it. The panel disingenuously says “Meta renews its Section 230 arguments on appeal,” which is because the Ninth Circuit requested supplemental briefings on 230).
* * *
The plaintiffs argued that Myanmar law, which doesn’t contain Section 230 immunity, should govern instead of US law. The panel rejects this argument:
The United States’ interest in applying Section 230 is clear. Imposing liability on Meta for its actions as a publisher would frustrate Section 230’s purpose of “promot[ing] the continued development of the Internet and other interactive computer services.”…
Myanmar’s interest in protecting its citizens from harmful attacks and misinformation on Facebook, while real, is insufficiently incorporated into the positive law of the country. Myanmar’s interest therefore does not predominate. For these reasons, even if we could or should consider Myanmar law, Section 230 applies
I am not a conflicts-of-laws expert, so I don’t know if the court’s methodology or conclusion is unusual. The fact that Section 230 protects a U.S. company being sued in the U.S. seems intuitive to me, even if the plaintiffs are foreigners and the offline harms took place in a foreign country. For more on Section 230’s extraterritorial application, see Prof. Chander’s paper.
In his self-concurrence (discussed further below), Judge Nelson says this part of the panel opinion (which he wrote) wasn’t necessary because Section 230 is a federal law, so it preempts any of California’s choice-of-law provisions. He has nine citations to the Federalist papers in this section, so he’s still living in the eighties (the 1780s).
* * *
The panel says that Section 230 applies to the plaintiffs’ claims, despite the plaintiffs’ invocation of the standard workaround arguments.
The panel says:
Plaintiffs characterize Meta’s duty as one of product design—that Meta should not have built Facebook in a way that boosted incitements to violence. Still, the alleged defects relate to Facebook’s core design as a publishing platform, particularly how Facebook promoted or downplayed third-party posts using algorithms. Under our case law, matching users with content is publishing conduct, even when the user has not requested the content. [cites to Doe v. Grindr, Dyroff, Carafano, Barnes]
This is a powerful statement: “matching users with content is publishing conduct.” This reinforces that Section 230 should not just apply to the substance of third-party content, but it should also apply to decisions about how to present that third-party content. It reminds us that “product design” choices by speech venues are synonymous with editorial decision-making, so “product design” claims against speech venues should implicate Section 230.
The plaintiffs tried several arguments to advance its design defects theories:
- “Meta did not do enough to screen and moderate content” => “Asking a platform
to monitor and review third-party content goes to Section 230’s core.” - Trying to fit into Doe v. Twitter’s “reporting infrastructure” exception, the plaintiffs highlighted that Facebook lacked a reporting function in the native language. => “Plaintiffs mentioned the alleged English-language reporting issues only once in their opening brief. Plaintiffs did not discuss the effect of Section 230 on their reporting infrastructure theory in their reply brief. Plaintiffs have forfeited any argument about their reporting infrastructure theory.”
- “Facebook’s promotion of posts is inextricably (even circularly) linked to the “social rewards” reflecting third-party engagement.” => “The encouragement provided by social rewards ultimately depends on third-party engagement and content. This dynamic is true even though Plaintiffs try to frame the issue as a matter of product design….Calling Facebook a “product” rather than a publication platform only obscures the point that Meta “published user-generated speech that was harmful” to Plaintiffs.”
A reminder: just because a service adopts a term with potential legal significance like “product” doesn’t mean the service has made an admission against its interest. See the Prager U v. YouTube case and its discussion of “public forum.”
The plaintiffs also tried the Anderson v. TikTok argument, i.e., the algorithm is Facebook’s first-party content. The court responds: “Facebook’s promotion of engagement-driven content through the 2009 recommendation algorithm must be characterized as recommending and matching conduct. We have already decided that such conduct is the work of publishers, rather than a platform’s own content or messaging.”
The panel also discusses the Roommates.com exceptions, saying:
The system of social rewards that Plaintiffs complain of is, “on [its] face, neutral.” There is no plausible allegation in the complaint that the Facebook algorithm specifically treated anti-Rohingya content differently than any other third-party content. What’s more, nothing in the complaint plausibly alleges that Meta singled out or selected violence-prone users to post violent content….
Even if the Facebook algorithm and system of third-party feedback and “social rewards” encouraged the posting of content, nothing about the platform’s design contributed to what made those posts illegal or actionable….
the algorithm does not change our calculus or count as a material contribution.
The plaintiffs also tried the decades-old argument that 230 only protects publishers, not distributors. Citing Calise, the panel replies:
This is a distinction without a difference. We have expanded Section 230 and “discarded the longstanding distinction between ‘publisher’ liability and ‘distributor’ liability” that existed at common law.
The panel concludes:
Meta created Facebook, an interactive computer service. That service recommended content to users to maximize engagement. Under our precedent, that is publishing conduct that does not materially contribute to or develop the underlying content.
Plaintiffs believe that Facebook’s design, coupled with the darker elements of human nature, caused real-world harm. But Section 230, as we have interpreted it, bars their claims, and we cannot hold Meta “responsible for the unfortunate realities of human nature.”
* * *
If the panel stopped there, the panel opinion would be what I consider to be a fairly straight-down-the-middle Section 230 opinion. The plaintiffs generally tried routine “product design” and “but the algorithms” arguments, which the Ninth Circuit has repeatedly rejected and rejects again. The case involves tragic facts, but the requested redress is outside the law’s boundaries. The ruling also casts significant doubt on the social media addiction rulings regarding Section 230 in several ways.
However…at the end of the panel opinion, the judges don’t stop talking. All three judges express support for en banc review of Section 230.
Judges Berzon and Fletcher’s Concurrence
This concurrence starts by saying that this ruling is a textbook Section 230 case:
We are bound by Ninth Circuit precedent addressing the scope of section 230 immunity, which requires the conclusion we reach here: The plaintiffs claims’ challenging Facebook’s algorithmic design seek to treat Meta “as the publisher or speaker of any information provided by another information content provider,” so section 230 provides Meta with immunity
However, these judges think there should be an algorithmic exception to Section 230:
I nevertheless continue to think that this Court’s precedent has unduly expanded the scope of section 230 immunity. For the reasons persuasively outlined by Judge Katzmann in his partial dissent in Force v. Facebook, 934 F.3d 53 (2d Cir. 2019), cert. denied, 140 S. Ct. 2761 (2020), and as I stated in my concurrence in Gonzalez v. Google LLC, 2 F.4th 871 (9th Cir. 2021), vacated, 598 U.S. 617 (2023), “if not bound by Circuit precedent I would hold that the term ‘publisher’ under section 230 reaches only traditional activities of publication and distribution—such as deciding whether to publish, withdraw, or alter content—and does not include activities that promote or recommend content or connect content users to each other
The logic flaw ought to be obvious: Judge Berzon believes the “traditional activities of publication and distribution…does not include activities that promote or recommend content or connect content users to each other.” But promoting and recommending content is one of the most central publication functions. Every publication decision is an act of prioritization–it elevates some content for greater attention over all other content that isn’t published. Further, book publishers don’t just publish books and then store them in a warehouse; they go out and promote the availability of the books so people will actually buy and read them.
Further, where Judge Berzon denigrates “connecting content users to each other,” she is using an unilluminating euphemism. Users aren’t just being “connected” to each other for shits and giggles. The connections facilitate the users talking to each other, i.e., AUTHORING AND DISSEMINATING CONTENT to each other. The speech venue’s role in making those connections–so people can communicate their content to each other–is a quintessential act of publishing content. Judge Berzon wants to kick out various activities from Section 230 by using an unnecessarily constricted and misguided definition of what she thinks constitutes “traditional activities of publication and distribution.”
This concurrence keeps digging. It embraces the logic defect that plagued the Anderson v. TikTok court: that Moody v. NetChoice (an opinion that didn’t discuss Section 230 at all) held that “an online platform’s decisions about how to order, organize, or display third-party content are the platform’s ‘expressive choices'” which, in turn, curtails Section 230. Thus:
Moody and Anderson persuasively suggest that algorithmic recommendations are an internet service provider’s first-party speech for purposes of assessing section 230 immunity
This concurrence says Section 230 doesn’t depend on how the algorithm is constructed:
generic, content-based algorithms do not bear any more resemblance to traditional publication than their “advanced”—individualized—counterparts….Whether generated by an “advanced,” “personalized recommendation” algorithm, an “engagement-maximizing” algorithm, or by hand, I understand suggestions and recommendations to be distinct messages presented by social media companies. Those messages cannot be attributed to any third parties and so are not protected by section 230.
I lament again how often people undervalue the curation function of publishers. The decision to publish or reject any individual item of content is obviously a curation function. It is similarly a curation decision to highlight some content items over others. Any efforts to distinguish between the two would inevitably shrink Section 230 down to, at most, an immunity for content hosting, i.e., Dropbox or Google Drive file hosting services. That was never Section 230’s intent, even in the simpler days of 1996. Moody v. NetChoice, which didn’t mention Section 230, did absolutely nothing to change the contours of Section 230.
These concurring judges summarize their position:
Whether soliciting scam advertisements or matching adults with minors on dating apps, this Court’s broad reading of section 230 of the Communications Decency Act permits internet service providers affirmatively to behave indecently, potentially—as alleged in Dryoff [sic] and Grindr—causing serious harm to vulnerable people. In my view, the conduct challenged in these cases stretches far beyond the traditional activities of publication contemplated by section 230
What does it mean for a publisher to “behave indecently”? I have no idea.
These judges conclude with a call for en banc hearing and maybe Supreme Court review:
there is now both an intercircuit conflict and strong tension with a recent Supreme Court case. Moody v. NetChoice and Anderson only underscore the necessity for en banc review of this Court’s precedent addressing section 230. I again—even more emphatically on this go round—urge this Court to reconsider en banc our precedent extending section 230 immunity to recommendation of content and connections to users
I still cannot wrap my head around how anyone agrees with the Anderson v. TikTok decision.
Judge Nelson’s Self-Concurrence
Judge Nelson writes yet another self-concurrence to a panel opinion he also wrote. This is not his first self-concurrence; and self-concurrences have become a bit of a status symbol/fetish among TAFS judges. I assume they think it signals that they are by-the-books adhering to the precedent in the main opinion but would have reached a different conclusion if they weren’t so restrained.
With respect to Section 230, he says “we have overread Section 230, straying from the original public meaning of the statutory text and creating an all-purpose liability shield for internet platforms.” (As I have repeatedly pointed out, Section 230 isn’t an “all-purpose liability shield”). For example:
how does a product liability claim hinge on treating the provider of an interactive computer service as the publisher of third-party content?…Even if a product liability claim might look like publishing third-party content from afar, the duty underlying such a claim is not “identical to publishing or speaking.”
I have no idea what Judge Nelson is talking about. Take, for example, publishing a field guide to mushrooms that misidentifies a poisonous mushroom as safe to eat. If the victim brings a products liability claim over the book’s mischaracterization, how is that claim not “identical” to imposing liability for publishing? What tendentious distinctions about the mechanics of publishing is Judge Nelson trying to make?
Judge Nelson also speculates that the products liability claims “likely fail” under California law because he is “skeptical” that Facebook is a “product.” First, it’s improper to speak specifically about this claim when it could possibly come back to the Ninth Circuit in this very case depending on how the case proceeds. Even worse, the claim wasn’t properly briefed for the Ninth Circuit by the parties.
Second, if he’s right that the claim should fail on its lack of merits, what exactly is the point of trying to fight to preserve the claim? The Ninth Circuit has already done a dozen of these “it’s not 230, but you still lose” switcheroos to plaintiffs, and yet it has not learned anything from that ever-growing process of wasted motion and futility. Judge Neslon says “we need not stretch Section 230 to reject” the products liability claims. In my view, his argument would instead nonsensically stretch Section 230 to preserve what he believes are futile products liability claims.
Judge Nelson also buys into the “but the algorithms” view: “Section 230’s protection of third-party publishing conduct should reach only “traditional” activities of publication and distribution—not every modern activity that bears some remote resemblance to it.” Except that because judges routinely undercount the curatorial functions of publishers, they are actually misperceiving the “traditional” activities of publishers.
To emphasize that Judge Nelson doesn’t understand what publishing means, he says:
Modern recommendation algorithms are opaque, esoteric, and—particularly when artificial intelligence enters the fray—incomprehensible, sometimes even to their own designers. Much of the matchmaking and network creation that modern algorithms engage in does not fit within any fair definition of publishing conduct
If an entity is gathering, organizing, and disseminating content, it is a publisher. Judge Nelson seems to think they are “qualifying” and “disqualifying” ways of organizing and disseminating content. It’s true that we’re now longer in the world of typesetting and ink-driven printing that prevailed during the writing of our Constitution. But the curatorial functions are indistinguishable. Surely Colonial printers would have enthusiastically used automated options if they had been available.
Judge Nelson also apparently agrees with the Anderson v. TikTok approach: “even if algorithms are publishing conduct, the Supreme Court has suggested that they are the publishing of the platforms’ own content…If an algorithm is a distinctive expressive product, then it makes sense that it is a platform’s own content, rather than the content of third parties…. it is unclear why algorithms should be immunized under Section 230 after Moody.” This is so confused. Curating third-party content is called “publishing,” and the curatorial choices don’t change the status of the third-party content as third-party content.
* * *
By not addressing the statute of limitations question at all, the panel leaves open an obvious problem. Assume the Ninth Circuit wipes Section 230 off the books entirely, for this case and all others. The case still lost on statute of limitations grounds, and the Ninth Circuit has never said that was incorrect. As a result, resolving Section 230’s applicability does nothing to resolve this case. The case remains dismissed on statute of limitations grounds until a court of greater authority says otherwise. That makes the panel’s engagement with Section 230 feel gratuitous. The concurrences are like advisory opinions to address some counterfactual hypothetical.
Thus, it was conspicuous how little Judge Berzon and Judge Nelson’s concurrences engaged with the facts of the Rohingya claims–or explain why this case doesn’t actually illustrate the problems they want to discuss. (Judge Nelson addresses this some–he thinks old Facebook algorithms aren’t the right test case for this pet theories). In most respects, the concerns in their concurrences are divorced completely from the case at hand. The case instead became a vehicle for venting their nonbinding views.
As you can imagine, the plaintiffs have already requested en banc review of this case. If the Ninth Circuit takes the case, it will set up a potentially existential battle over what it means to “publish” content, the role and value of publisher curation, and how a Supreme Court pro-free speech opinion that had nothing to do with Section 230 might nevertheless curtail online speech. Expect some turbulent jurisprudential times ahead.
Case Citation: Doe 1 v. Meta Platforms Inc., 2026 WL 1144707 (9th Cir. April 28, 2026)