The First Amendment to the US Constitution safeguards “freedom of speech” from government censorship. But does it––and should it––protect speech created by modern technological tools that involve minimal human effort, like the simple push of a digital button or the entry of a short query? Put differently, does de minimis mortal exertion in generating meaningful (to some) expression cramp constitutional coverage for such technologically derived output?
Answering these questions is vital because new technologies––generative artificial intelligence tools––reduce the amount of mental and physical human effort needed to produce expression. The issue, however, predates generative AI.
In 2012, a Virginia federal district court concluded in Bland v. Roberts “that merely ‘liking’ a Facebook page is insufficient speech to merit constitutional protection.” Judge Raymond A. Jackson reasoned that “[s]imply liking a Facebook page . . . is not the kind of substantive statement that has previously warranted constitutional protection.” The case involved a sheriff’s employee who claimed he was retaliated against after liking a post from a rival candidate running for sheriff.
The US Court of Appeals for the Fourth Circuit, however, reversed that ruling in 2013. It observed that “clicking on the ‘like’ button literally causes to be published the statement that the User ‘likes’ something, which is itself a substantive statement.” Critically, the miniscule effort of producing the thumbs-up like symbol was irrelevant for the Fourth Circuit: “That a user may use a single mouse click to produce that message that he likes the page instead of typing the same message with several individual key strokes is of no constitutional significance.”
Courts have grappled with a similar issue when it comes to using smartphone technology to capture videos of police doing their jobs in public places. Is the rudimentary conduct of pushing a record button to create a video recording protected by the First Amendment? The US Court of Appeals for the Seventh Circuit addressed that issue in 2012 in American Civil Liberties Union of Illinois v. Alvarez. It determined that “[t]he act of making an audio or audiovisual recording is necessarily included within the First Amendment’s guarantee of speech and press rights as a corollary of the right to disseminate the resulting recording. The right to publish or broadcast an audio or audiovisual recording would be insecure, or largely ineffective, if the antecedent act of making the recording is wholly unprotected.” (emphasis in original).
That logic may see like common sense, but other courts have disagreed. To wit, a Pennsylvania federal district court in 2016 in Fields v. City of Philadelphia considered “whether photographing or filming police on our portable devices without challenging police is expressive conduct protected by the First Amendment.” Judge Mark A. Kearney concluded that the act of silently recording police without criticizing them did not amount to expressive conduct and thus did not merit constitutional protection. In short, “non-confrontational ‘observing’ and ‘recording’” didn’t amount to protected expression.
The US Court of Appeals for the Third Circuit, however, reversed Kearney’s ruling in 2017. It squarely held that “the First Amendment protects the act of photographing, filming, or otherwise recording police officers conducting their official duties in public.” It reasoned that “[t]he First Amendment protects actual photos, videos, and recordings . . . and for this protection to have meaning the Amendment must also protect the act of creating that material.” In 2022, the Tenth Circuit affirmed this principle, concluding that “video recording is ‘unambiguously’ speech-creation, not mere conduct.”
The value of content generated by the trifling human effort in Bland v. Roberts (the Facebook liking case) and in the recording-police cases explains why First Amendment protection is essential for the act of creating it. Bland involved core political expression (supporting a candidate for public office) affecting democratic self-governance (the voting of wise decisions, as Alexander Meiklejohn would’ve put it). The act of liking also serves the First Amendment value of “individual self-realization,” allowing a person to identify himself as supporting a candidate. Recording cops, in turn, might expose “police abusing their authority” while simultaneously protecting them from false and malicious claims of wrongdoing.
Should the First Amendment protect original output created by generative AI in response to nominal human efforts in entering prompts? Scholars Eugene Volokh, Mark Lemley, and Peter Henderson answer “yes” in a 2023 article. Generative AI, they assert, simply is “a special case of the proposition that the First Amendment protects technologies that make it easier to speak.” It not only helps users speak (say, by drafting letters), but affects the First Amendment “interests of listeners in receiving meaningful communication” about important issues. Both the right to speak via generative AI and the right to receive speech justify sheltering speech creation and output.
In sum, minimal effort in creating speech via new technologies should not detract from constitutional protection.
The post Minimal Human Effort and New Speech-Creating Technologies: Protecting Both Conduct and Content appeared first on American Enterprise Institute – AEI.