Machine Speech: Towards a Unified Doctrine of Attribution and Control

Authors

Brian Sites, University of Miami School of Law

Abstract

Like many courts across the country in 2023, courts in the Eleventh Circuit were met with novel claims challenging ChatGPT and other artificial intelligence tools. These cases raise common questions: How should courts treat the speech of machines? When a machine generates allegedly defamatory material, who is the speaker—mortal or machine? When a machine generates expressive creations, who is the artist, and does that shape copyright eligibility? When a machine makes assertions about reality through lab analyses and other forensic reports, who is the accuser, and how does the answer impact a defendant’s rights at trial? Should those answers stem from a common doctrinal model for machine speech attribution, or should they be discordant based on competing policy goals in each area of law? This Article explores those cases and the increasingly expansive machine speech doctrine, including Walters v. OpenAI, a 2023 Georgia defamation case involving ChatGPT, and a series of cases arising from the Eleventh Circuit’s machine speech Confrontation Clause case—one of the first major cases of its kind—United States v. Lamons.

Recommended Citation

Brian Sites, Machine Speech: Towards a Unified Doctrine of Attribution and Control, 78 U. Mia. L. Rev. 1025 (2024)
Available at: https://repository.law.miami.edu/umlr/vol78/iss4/4