The Brandywine Review

Commentary on American law

Media and Speech Law

Media and Speech Law

The federal government has one statute on the books protecting a customer's right to post a harsh online review, and no statute at all protecting the person being sued over having posted one. That gap is where most of the live litigation in this area actually sits.

The Consumer Review Fairness Act, 15 U.S.C. § 45b, voids, from the moment they are signed, form contract provisions that bar a customer from posting a review, penalize her for posting one, or force her to sign over the intellectual property rights in what she writes. A violation is treated as a per se unfair or deceptive practice, and the Federal Trade Commission or a state attorney general can bring an enforcement action. On paper the statute is a categorical rule with real teeth. The agency's own enforcement record under the statute by itself is thin: three settled actions against small businesses in May 2019, two more in June, and no comparable wave since. A company sued by an unhappy customer for exercising a right the CRFA plainly gives her is, in most cases, not defended in that lawsuit by the CRFA at all. It is defended, when it is defended successfully, by whatever anti-SLAPP statute the state where she posted the review happens to have.

Roughly forty states and the District of Columbia now have some form of anti-SLAPP law, and the strongest of them, including California's and New York's statute as strengthened in 2020, reach any communication in a public forum on a matter of public concern rather than only petitions to government, broadly enough to cover an ordinary consumer review of a contractor or a landlord. A defendant who can invoke one of these statutes gets an early motion that shifts the burden onto the plaintiff to show a likelihood of success before discovery grinds on, and often collects attorney's fees if the motion succeeds. A defendant sued in a state without a comparable statute has no equivalent tool and has to survive to summary judgment the ordinary way, and a defendant sued in federal court sitting in diversity fares no better or worse than that depending on which circuit hears the case, since the First and Ninth Circuits let a state's anti-SLAPP statute apply in federal court while other circuits do not.

That second qualification carries more weight than it looks like it should, because federal circuits disagree sharply about whether a state anti-SLAPP statute even applies in federal court in the first place.

Abbas v. Foreign Policy Group, LLC, 783 F.3d 1328 (D.C. Cir. 2015), held that the District of Columbia's anti-SLAPP special motion to dismiss cannot be used in federal court, because it answers the same question, whether a complaint survives a defendant's early motion to end the case, that Federal Rules of Civil Procedure 12 and 56 already answer, and under the Erie doctrine a state procedural rule cannot displace a Federal Rule doing the same job. Then-Judge Kavanaugh wrote the opinion. Klocke v. Watson, 936 F.3d 240 (5th Cir. 2019), reached the identical conclusion about the Texas Citizens Participation Act, resolving a question the circuit had repeatedly declined to answer in earlier appeals rather than reversing any settled practice of its own; the panel found itself unbound by a decade-old Fifth Circuit case applying Louisiana's anti-SLAPP statute, reasoning that Louisiana's law imposed lighter burdens than Texas's and that the earlier case never analyzed the conflict with the Federal Rules at all. The Ninth Circuit has gone the other way for decades, treating state anti-SLAPP statutes as substantive enough to survive alongside the Federal Rules rather than conflict with them. A defamation defendant with an identical set of facts can be entitled to an early fee-shifting motion in one federal courthouse and denied any such motion in another, purely because of which circuit's Erie analysis controls, with nothing about the underlying speech or the underlying state statute any different between the two cases.

Right of publicity claims sit on the other side of the same First Amendment line, protecting a person's commercial interest in her own identity rather than a speaker's interest in criticizing someone else, and the doctrine built to balance that interest against expression is now being asked to do work it was never designed for.

Comedy III Productions, Inc. v. Gary Saderup, Inc., 25 Cal. 4th 387 (2001), gave California, and by extension the other courts that have since looked to its reasoning, a transformative use test: a work that adds significant new expression, meaning, or message to a person's likeness earns First Amendment protection against a right of publicity claim, while one that merely trades on a literal, marketable depiction of the person does not. Saderup's charcoal drawings of the Three Stooges lost under that test because the California Supreme Court found no significant transformative content, only skilled reproduction aimed at commercial exploitation of the image itself.

A new section of New York's right of publicity law, enacted in November 2020 and effective the following May, codified at N.Y. Civil Rights Law § 50-f, takes a different approach to a problem Saderup's era did not have to consider: a digital replica, defined as a highly realistic, computer-generated image or voice that either never involved the real person or materially altered what she actually did. The statute imposes liability for using a deceased performer's digital replica without consent, but carves out exceptions for parody, satire, commentary, and criticism, the same categories a transformative use analysis would ask about anyway. The two frameworks do not obviously fit together. A deepfake realistic enough to trigger liability under section 50-f's own definition is, by that same definition, not transformative in Saderup's sense; it mimics rather than reinterprets. That leaves the parody exception to do all the work transformative use analysis used to do on its own, and neither the statute nor any court applying it has yet said how a genuinely transformative deepfake, one that alters meaning rather than merely copying a voice, is supposed to be told apart from the kind the statute was written to stop.