The Volokh Conspiracy
Mostly law professors | Sometimes contrarian | Often libertarian | Always independent
The First Amendment and Restrictions on Gathering Information by Drone
From Sixth Circuit Judge John Bush's very interesting opinion Friday respecting denial of rehearing en banc in Yoder v. Bowen:
This case involves an as-applied challenge to a Michigan law (the drone statute) that makes it illegal to "us[e] an unmanned vehicle or unmanned device that uses aerodynamic forces to achieve flight"—i.e., a drone—while "tak[ing] game or fish." Drone Deer Recovery, a plaintiff here, offers a service where it tracks downed animals using drones and then posts the location of the animals' carcasses online so hunters can more easily find their kill. The plaintiffs allege that the drone statute violates their First Amendment rights because it (1) is a content-based speech restriction, (2) violates the speech-inputs doctrine, and (3) unconstitutionally restricts their ability to engage in inherently expressive conduct. The panel rejected all three arguments and determined that the statute survived intermediate scrutiny.
{We have referred to a type of protected speech as "speech inputs," but it goes by different names in different jurisdictions. The Tenth Circuit, for example, has referred to it as "the protected creation of speech," and the Supreme Court has referred to it as "[s]peech in aid of" protected speech.}
I write separately because I have concerns about the panel's reasoning related to the speech-inputs doctrine. The Supreme Court has indicated that "heightened scrutiny"— something more than O'Brien intermediate scrutiny—applies when the government seeks to ban the means to create speech. See Sorrell v. IMS Health Inc. (2011) (noting that a Vermont statute banning the sale of certain pharmaceutical data effectively banned certain entities from speaking with physicians and pharmaceutical companies and was therefore subject to "heightened scrutiny").
For example, if a State enacted a statute banning the ownership of pens and paper, the statute would likely violate the First Amendment under the speech-inputs doctrine because it would restrict the ability to express thoughts through handwriting. Here, drone-obtained information may be analogous to pens and paper because it provides what the plaintiffs allege is a critical input needed for Drone Deer Recovery's speech to hunters. But the panel declined to apply Sorrell's more rigorous level of review.
The panel's error may be understandable given the confused state of the speech-inputs doctrine following Sorrell. After all, that case is far from a model of clarity. The words "heightened scrutiny" have sometimes been considered synonymous with "intermediate scrutiny." But Sorrell also tells us that the statute at issue in that case "enact[ed] content- and speaker-based restrictions on the sale, disclosure, and use of prescriber-identifying information." Content-based speech restrictions are normally subject to strict scrutiny. Meanwhile, Sorrell did not find the statute's content-based speech restriction to be dispositive and proceeds to apply Central Hudson's commercial speech test, which is an entirely different inquiry from strict or intermediate scrutiny….
Sorrell is also unclear in the degree to which a speech input needs to be restricted before the doctrine comes into play. On the one hand, it seems like banning all pens and paper would easily violate the doctrine because that would outlaw the handwritten word. By contrast, a restriction on the use of a specific chemical in printer ink might not because printer ink still remains readily available. But Sorrell does not give us any direction on how to distinguish between the two types of regulation.
And, as a third point of confusion, the Supreme Court has never clarified how exclusively dedicated to creating speech the input must be before it receives some level of scrutiny under the First Amendment. When technology may be employed for purposes other than the generation of speech, those other uses perhaps may attenuate the level of protection for speech associated with use of the technology. Is a drone a speech input? Is a microchip inside the drone that is vital to its functioning?
Sorrell leaves more questions than answers, and the panel only added to the confusion. The panel determined that drones are not speech inputs, and then applied intermediate scrutiny anyway. But if the drones are not speech inputs, then it's unclear why any level of scrutiny would apply
The panel seeks to distinguish this case from speech-inputs precedents because Drone Deer Recovery's speech is not political. But the alleged wrong from prohibiting drone usage does not depend on whether the speech is political. Rather, the constitutional violation from banning a speech input arises when the restriction effectively abolishes the speech altogether. It is one thing to say that business-related speech, once expressed, may have less First Amendment protection than political speech; it is quite another thing to deprive the speaker of the means for expressing its speech in the first place….
In Sorrell, the Supreme Court held that a Vermont statute "restrict[ing] the sale, disclosure, and use of pharmacy records that reveal the prescribing practices of individual doctors" was unconstitutional because it "imposed a restriction on access to information" that could be used "in aid of pharmaceutical marketing"—i.e., it limited access to a critical speech input for pharmaceutical marketing. That case involved a statute completely divorced from politics, yet the Court still applied the speech-inputs doctrine…. So the panel was mistaken to the extent that it sought to distinguish Sorrell based on the non-political nature of Drone Deer Recovery's speech.
{As an aside, I struggle to see how the panel's distinction between political speech and non-political speech would matter here, given that the drone statute seems to regulate political speech as well. The drone statute would appear to apply equally to political speech. For example, the law would also foil the speech of animal rights activists who want to track down felled game and use it to protest animal cruelty—a quintessential form of political speech.}
The panel also attempted to distinguish Sorrell because the law in that case was not content neutral, given that the Sorrell statute allowed prescribing information to be used for some purposes but not others. But the drone statute similarly restricts a speech input based on the content of its use. The law forbids employing drones to obtain and deliver the location of felled game. But the law allows drones to deliver any other kind of information. For example, the statute apparently does not prohibit using drones to obtain and deliver data about the number and types of trees, the location of trails, etc. The drone statute thus regulates the speech input based on the content of speech for which the information will be employed. This is a content-based regulation much like in Sorrell.
We thus must apply Sorrell to this case. But what is the standard that Sorrell requires us to apply? I believe, based on the speech-inputs doctrine, it may be a higher level of review than the panel applied.
The panel's rejection of the speech-inputs doctrine may portend a split between our circuit and the Fourth and Ninth Circuits. In an opinion that came down after the plaintiffs petitioned for rehearing en banc, the Ninth Circuit concluded that an ordinance banning observing sideshows (a form of reckless driving in an intersection) was unconstitutional because it inhibits "the process of creating a form of pure speech." Garcia v. Cnty. of Alameda (9th Cir. Sept. 4, 2025). The court explained that, even though observing a sideshow might be a restriction on conduct, it was entitled to First Amendment protection because it regulated "a predicate for … recording of those events," meaning that it essentially outlawed a speech-input. {The Supreme Court has held that newsgathering is protected under the First Amendment.} Meanwhile, the Fourth Circuit has held that a statute banning organizations from planting moles (i.e., undercover spies) in farms and slaughterhouses bans a speech input because it "prevents an undercover employee from publishing a critical article based on any notes she takes of documents or policies laid out in a breakroom."
The statutes in Garcia and PETA are a bit afield of the facts of this case, but the panel's reasoning may be difficult to square with those cases. If observing employees in a slaughterhouse or watching reckless drivers in an intersection are speech inputs governed by Sorrell, then it would appear that observing animals via a drone would also be such a speech input. To be sure, the panel might say PETA involved political speech because the plaintiff was an animal rights advocacy group. But the plaintiff in Garcia was a transportation reporter, and there is no indication from that case that he was engaged in political speech.
If this case began and ended as a hunting-with-drones precedent, it perhaps would not be worth delving so deeply into the panel's rationale for its decision. But I worry that the panel's opinion may be interpreted to diminish First Amendment protection more broadly, including for academics and journalists.
Consider how the panel's reasoning could be employed to diminish academic freedom. Many academic studies rely on recorded interviews. A State could theoretically enact a statute banning the recording of interviews between a healthcare provider and a patient. This prohibition would make it unlawful, for example, to use recorded interviews in a study examining whether psychologists can convince children to remember traumatic events that did not happen. Such a study could radically change the way we consider witness testimony in many criminal trials. And yet, employing reasoning similar to the panel's rationale here, a State's transparent attempt to stifle that research could be subject to mere O'Brien intermediate scrutiny because (1) a psychology paper is not political speech, and (2) the statute only restricts employment of a particular technology (a recording device) that can be analogized to the drone usage in this case.
The potential effects of the panel's reasoning could be similarly problematic for journalists. Consider a statute that bans audio or video recorded interviews altogether. If a journalist wanted to document, for example, eyewitness accounts of athletes who gambled on their own games, this statute would effectively ban that form of journalism that relates to a non-political topic. And yet, even though journalism (muckraking in particular) is one of the First Amendment's central concerns, a statute restricting these journalists' recordings would be subject only to O'Brien intermediate scrutiny simply because (1) the interview does not involve political speech and (2) the law did not ban the interview itself but only particular ways of recording the interview. Using the rationale advanced to defend the drone statute—that the drone statute only bans a particular technology to gather information but leaves in place traditional methods for tracking killed prey—one could argue that banning video and audio recordings of interviews is acceptable because the journalist can still use the traditional pen-and-paper method to memorialize those interviews.
These two hypotheticals cover academic research and journalism—areas that are supposed to receive the highest levels of First Amendment protection, even when they do not implicate political speech. See, e.g., Fla. Star v. B.J.F., 491 U.S. 524, 541 (1989) (applying strict scrutiny to statute limiting a journalist's ability to publish the name of a sexual assault victim); Keyishian v. Bd. of Regents (1967) ("Our Nation is deeply committed to safeguarding academic freedom, which is of transcendent value to all of us and not merely to the teachers concerned. That freedom is therefore a special concern of the First Amendment."). And yet, under the panel's reasoning, they may receive only O'Brien intermediate scrutiny—with significant consequences, indeed.
That said, although I have concerns about the panel's opinion, I do not think that this case is a viable candidate for rehearing en banc. The panel's reasoning stems from a difficult-to-interpret Supreme Court opinion, and we are powerless to modify the directives from a controlling Supreme Court opinion in any way. So all that is left to do is wait for further guidance from the Supreme Court.
I cannot blame the panel opinion for its attempt to sort through confusing Supreme Court precedent. After all, Sorrell appears to call for O'Brien intermediate scrutiny, strict scrutiny, and Central Hudson scrutiny, all at the same time. Thus, three people could theoretically argue in favor of each separate standard of review, and Sorrell would provide equally strong support for each position. But I am still concerned that the panel's reasoning in this case might cause problems down the road. Ultimately, I hope that the Supreme Court will give plenary consideration to this case or one like it to clarify the parameters of the speech-inputs doctrine.
