Judicial Approaches to Acknowledged and Unacknowledged AI-Generated Evidence
Science and Technology Law Review, Vol. 26, No. 2 (2025)
A framework for two evidentiary problems that look similar but are not: evidence everyone agrees was produced by AI, and evidence one side alleges is a deepfake.
Overview
The article distinguishes acknowledged AI-generated evidence from unacknowledged AI-generated evidence. The first category raises questions about how an AI system produced its output and whether that output is sufficiently reliable and helpful. The second raises a threshold authenticity problem: one side presents an image, recording, text, or video as real, while the other claims it was fabricated. The authors examine how existing evidence rules apply to each problem and where new rules may be needed.
Key contributions
- Separates acknowledged AI output from alleged but unacknowledged synthetic evidence.
- Reviews the limits of human and automated deepfake detection.
- Analyzes authenticity, relevance, expert testimony, and unfair-prejudice issues under existing evidence rules.
- Describes two proposed new rules submitted for consideration by the Advisory Committee on Evidence Rules.
- Ends with practical preparation steps for judges and lawyers.
Abstract
Between 2014 and 2024, rapid advancements in computer science ushered in a dramatic new form of technology—Generative AI (“GenAI”). It offered seemingly limitless possibilities for creative applications never before imagined. But it also brought with it a darker side—the ability to create synthetic or “fake” text, images, audio, and audiovisual depictions so realistic that it has become nearly impossible—even for computer scientists—to tell authentic from fake content. The article explores the development of GenAI and the deepfake phenomenon; distinguishes acknowledged from unacknowledged AI-generated evidence; considers the flexible application or modification of existing evidence rules; describes two proposed new rules; and identifies practical steps for judges and lawyers.
Citation
Maura Grossman & Paul Grimm, Judicial Approaches to Acknowledged and Unacknowledged AI-Generated Evidence, 26 Science and Technology Law Review, no. 2 (2025). https://doi.org/10.52214/stlr.v26i2.13890