Let’s Get Real: Weak Artificial Intelligence Has Free Speech Rights

By James B. Garvey

The right to free speech is a strongly protected constitutional right under the First Amendment to the U.S. Constitution.  In 2010, the U.S. Supreme Court significantly expanded free speech protections for corporations in Citizens United v. FEC.  This case prompted the question:  could other nonhuman actors also be eligible for free speech protection under the First Amendment?  This inquiry is no longer a mere intellectual exercise:  sophisticated artificial intelligence (AI) may soon be capable of producing speech.  As such, there are novel and complex questions surrounding the application of the First Amendment to AI.  Some commentators argue that AI should be granted free speech rights because AI speech may soon be sufficiently comparable to human speech.  Others disagree and argue that First Amendment rights should not be extended to AI because there are traits in human speech that AI speech could not replicate.

This Note explores the application of First Amendment jurisprudence to AI.  Introducing relevant philosophical literature, this Note examines theories of human intelligence and decision-making in order to better understand the process that humans use to produce speech, and whether AI produces speech in a similar manner.  In light of the legal and philosophical literature, as well as the Supreme Court’s current First Amendment jurisprudence, this Note proposes that some types of AI are eligible for free speech protection under the First Amendment.