Third Circuit Says Section 230 Doesn’t Apply to Publicity Rights Claims–Hepp v. Facebook

The Third Circuit ruled today that Section 230 doesn’t preempt publicity rights claims because they qualify as “intellectual property” claims. This ruling directly conflicts with the Ninth Circuit’s rule, which says that all state IP claims are preempted by Section 230. This creates the foundation for a possible Supreme Court review of this issue.

About the Case

Hepp is a television newscaster. At issue is a photo taken in public without Hepp’s consent that got cycled into various memes and unwanted commentary. Hepp sued a variety of entities hosting variations of the photos, but the court dismisses several defendants for lack of jurisdiction. That leaves only a single version of the photo at issue: a dating app, FirstMet, used the photo for an ad that it ran on Facebook. I’m not sure if Hepp has sued FirstMet, but they aren’t in this appeal. Instead, the only defendant remaining is Facebook, and the only question on the table is whether Facebook can preempt Hepp’s publicity rights claims against it due to Section 230. The district court, in a surprise decision, said yes. In a less surprising decision, the Third Circuit (in a split opinion) says no.

Legal Background

Section 230(c)(1) says websites aren’t liable for third-party content. Section 230(e)(2) says “Nothing in this section shall be construed to limit or expand any law pertaining to intellectual property,” so IP lawsuits over third-party content are not preempted by Section 230.

In 2007, the Ninth Circuit in Perfect 10 v. ccBill held that “intellectual property” meant only “federal intellectual property” because otherwise this exception would expose Internet services to heterogeneous state laws that they couldn’t manage. While this may sound a little wacky, it makes sense that Section 230 was designed to ensure that Internet services only need to comply with a single national standard as much as possible. That’s why, for example, Section 230 preempts state criminal prosecutions (except for the recent FOSTA exception) but preserves federal criminal prosecutions–that approach ensures that Internet services have only a single body of criminal law to worry about.

Based on the ccBill precedent, a number of Ninth Circuit-governed courts have applied Section 230 to publicity rights claims. See, e.g., Ripple v. YouTube.

Despite this, a textualist reading of Section 230(e)(2) may point in a different direction. The law says “intellectual property,” not “federal intellectual property,” and adding that word to the statute via the common law has a pretty significant substantive effect. Not surprisingly, courts outside the Ninth Circuit have almost uniformly disagreed with the ccBill ruling. That made the district court ruling in this case noteworthy. It was a full-bodied embrace of the Ninth Circuit interpretation, which created the possibility of the Ninth Circuit’s rationale expanding nationwide.

The Majority Opinion

The majority rejects Facebook’s Section 230 preempts arguments both on statutory analysis and policy grounds.

With respect to statutory analysis, the majority says “when Congress wanted to cabin the interpretation about state law, it knew how to do so—and did so explicitly.”

With respect to policy, the majority goes property-absolutist: “Because state property rights can facilitate market exchange, interpreting the § 230(e)(2) limitation to include state intellectual property laws tracks Congress’s pro-free-market goal.” Not only has it been proven a zillion times that property rights often harm free markets, but this concern doesn’t fit Hepp’s situation at all. The majority says “if likeness interests are disregarded on the internet, the incentives to build an excellent commercial reputation for endorsements may diminish…information provided by promotional advertisements can enhance market efficiency and vibrancy” but these are orthogonal statements. (I even agree with the latter point!) If FirstMet violated Hepp’s publicity rights, they should be accountable for it. That’s not the question before this court.

With respect to the concerns that state publicity rights laws are a doctrinal mess, the court says that the Lycos and Project Playlist rulings have been around a long time and “yet neither Facebook nor its amici provide evidence that the rulings created the disarray they now predict.” I’m not sure about where the majority formed this empirical conclusion. In fact, Internet publicity rights/230 cases are a jurisprudential disaster (check out some confused rulings in this category).

The majority says that Hepp’s statutory publicity rights claim qualifies as an intellectual property claim. The court cites several dictionaries that define publicity rights as an intellectual property (including Black’s Law Dictionary), but other dictionaries don’t, so the court holds a weird bake-off to assess the zeitgeist of dictionaries.

This leads to a twisted conclusion that dictionaries include trademarks as IP (not surprisingly), so “that inclusion implies the right to publicity by analogy [because] the right of publicity and trademark are close analogues.” Later, the court explains:

Just as a counterfeit item can misappropriate a trademark owner’s goodwill, so too might the unauthorized use of Hepp’s image in the ad. Further, both misappropriations could create consumer confusion and undercut efficient incentives. In this way, trademark claims typically avoid violating free speech by addressing misleading commercial speech. So too for Hepp’s statutory claim against Facebook. Thus, Hepp’s statutory claim against Facebook is about the commercial effect on her intellectual property, not about protected speech.

Ugh, no. Trademarks come from consumer protection roots while publicity rights come from privacy roots, so publicity rights usually lack critical doctrinal features of trademark law like a “use in commerce” prerequisite for protection (but see the discussion below about Pennsylvania’s law) and a likelihood of consumer confusion for violations. Thus, collapsing the trademark/publicity rights distinction is not an intellectually defensible position. As the minority opinion says in a footnote, “Although the majority compares the right of publicity to patent, copyright, and (especially) trademark rights, they clearly are not the same thing.”

(The minority opinion, in a different footnote, has another good zinger about the fact that the Ninth Circuit said Section 230 applies to Lanham Act false advertising claims, even though they are literally housed in the same section as federal trademark law, so being close to trademark law doesn’t confer a free pass around Section 230).

The majority ends with a not-credible declaration that its ruling doesn’t threaten free speech. The court says this by referring to the requirement in Pennsylvania’s publicity rights statute that it only protects people “whose valuable interest in their likeness ‘is developed through the investment of time, effort, and money.'” The court bolsters that with a disingenuous statement that it’s not opining about publicity rights laws without that requirement, though everyone will read this case to apply to those claims as well. And irrespective of the preconditions for Pennsylvania publicity rights protection, as Jessica Silbey and I documented in our Copyright’s Memory Hole paper, plaintiffs pile into the IP exception to Section 230 for the express purpose of chilling free speech. This opinion reinforces–indeed, encourages–those efforts.

The Minority Opinion

The minority summarizes:

the 2007 Perfect 10 decision was the only circuit court clearly on point, and it kept the proverbial door closed on a potential influx of disparate and downright confusing state law “intellectual property” claims that would be contrary to Congress’s express goals in enacting § 230. We now open this door and, as I explain in more detail below, this drastic step undermines the broad policy objectives codified in § 230.

Regarding the statutory analysis, the minority asks: “Would Congress have really gone so far as to grant immunity from a wide range of state and federal laws—including state criminal law—yet permit claims to go forward under the nebulous (and expansive) category of state “intellectual property”/“rights of publicity” laws?” Regarding the policy considerations, the minority opinion recounts some of the anarchy that plagues publicity rights laws nationwide; and correctly says “[t]hese differences are not surprising given the confusion that exists regarding the specific basis of the right of publicity.”


Future Proceedings. This case is a good candidate for en banc review. The majority opinion made two major errors. First, it took a property rights absolutist position, something that is misguided when it comes to publicity rights, which necessarily must engage in a more balanced/nuanced approach given their intersections with First Amendment-protected speech. Second, the majority’s overly generous equation of publicity rights to trademark harms both Section 230 jurisprudence and possibly many other areas of law. The majority’s approach was undertheorized and pernicious–exactly the kinds of things that benefit from en banc review.

If the opinion doesn’t get fixed in the Third Circuit, the unambiguous conflict with the Ninth Circuit sets up a circuit split that could support Supreme Court review. I’m dreading when any Section 230 case goes to the Supreme Court because Justice Thomas will feel free to mess up Section 230 jurisprudence, even on topics not properly before the court. So I can’t say I’m eager to see this case get there, but this case would be better positioned to get cert than many other Section 230 cases.

“Intellectual Property” Is Inherently Ambiguous. When Congress enacted the Defend Trade Secrets Act (DTSA), it specifically said that the new trade secret rights created in the law are not “intellectual property” claims. The whole point of doing so was to preserve the existing Perfect 10 v. ccBill rule that trade secret claims were covered by Section 230. That’s a pretty powerful Congressional endorsement of the ccBill rule, a point made by the minority in a footnote but ignored by the majority.

In my analysis of that DTSA provision, I researched how federal law defines “intellectual property.” The bottom line: it’s a mess. Indeed, the term “intellectual property” has always been ambiguous. I know we have norms about what should be covered in our Intellectual Property survey courses–which almost always include the big four of trade secrets, copyrights, trademarks, and patents–but there are dozens or hundreds of other laws that protect intangible assets that could also be considered “intellectual property” or “para-IP” rights, and we don’t have a principled way to decide which of those laws are in or out of scope. The majority’s opinion doesn’t seek to resolve that scope problem. If anything, the majority opinion adds to the confusion with its weird near-equation of trademarks with publicity rights.

(There is a more structural objection to “intellectual property” that the term “property” should not be used for intangible assets at all. A different descriptor might have sidestepped this problem entirely. My point is that the term was facially unclear when adopted and remains unclear today).

Hepp Should Still Lose on Remand. Hepp should still lose her case on the merits. Facebook published a third-party ad. I believe Pennsylvania’s statute required notice-and-takedown to impose publisher liability: “No person, firm or corporation, including their employees and agents, in the business of producing, manufacturing, publishing or disseminating material for commercial or advertising purposes by any communications medium shall be held liable under this section unless they had actual knowledge of the unauthorized use of the name or likeness of a natural person as prohibited by this section.” Indeed, publicity rights claims imposed on publishers get heightened First Amendment scrutiny. A notice-and takedown approach may be constitutionally required. I don’t recall if Hepp ever sent a takedown notice to Facebook and what happened after that. If Hepp never sent a takedown notice, or if Facebook promptly honored it, I think Facebook should win.

Case citation: Hepp v. Facebook, 2021 WL 4314426 (3d Cir. Sept. 23, 2021)