This article first appeared on Law360 on June 12, 2025.
As AI and machine learning tools become more common in expert workflows, courts are increasingly scrutinizing their use under Federal Rule of Evidence 702. In The Risks and a Path Forward for AI and Machine Learning Tools in Expert Testimony, we examine recent cases where AI-assisted opinions were excluded due to concerns about reliability, transparency, and expert understanding.
Published in Law360, the article draws parallels to the courtroom journey of statistical tools like SAS and Stata, once novel and now widely accepted thanks to transparent methods and qualified expert use. This evolution offers a roadmap for how AI and ML might achieve similar acceptance.
Key recommendations include:
- Verifying AI-generated data
- Disclosing AI use and methods
- Ensuring expert understanding of tools
- Demanding transparency from vendors
Used responsibly, AI tools can support expert testimony but only if they meet the same standards of reliability and transparency that courts expect from established methodologies.