Evidence
AI has introduced a broad range of complex evidentiary questions for judges. For example, how do judges assess the validity and reliability of acknowledged and unacknowledged GenAI evidence, disputes over deepfakes, and the impact of GenAI evidence on juries. This section offers an introduction to some of the frequently occurring evidentiary challenges and includes links to more in-depth analysis.
-
Acknowledged AI-Generated
Examples of 'acknowledged AI-generated evidence' include the use of GenAI by an employer to screen resumes, by law enforcement to enhance low-quality surveillance video, and by an expert to prepare a written report. Challenges to the admissibility of this evidence often present difficult technical questions for the court to consider.
-
Unacknowledged AI-Generated (Deepfakes)
Unacknowledged AI-generated evidence (often referred to as ‘deepfakes’) is evidence that the proffering party represents as authentic but the challenging party claims is AI-generated or manipulated. When evidence is alleged to be a deepfake, judges may be called upon to assess its admissibility.