Insights / AI in Personal Injury Litigation
The Future of Personal Injury Litigation Is AI-Assisted, Not AI-Dependent
By Breen LNC.
Short answer: What we can verify points to AI assisting people, with people still answerable. A federal court has sanctioned unchecked output, rule makers are studying machine evidence offered without an expert, and clinical studies suggest AI drafts with human review may be viable. A firm that depends on AI alone carries risk it cannot hand off.
The courtroom is not ready to take a machine’s word
Expert opinion in federal court runs through Rule 702, which requires the proponent to show the court it is “more likely than not” that the testimony rests on sufficient facts and reliable methods, reliably applied. Rule makers have recently asked what should happen when a machine’s output is offered without an expert.
A proposed Rule 707 would have applied Rule 702’s reliability requirements to machine-generated evidence offered without an expert. After public comment, the Advisory Committee on Evidence Rules reported in May 2026 that it did not recommend action on the proposal at that time. It revised the draft and planned further study. Rule 707 is not law, and the text may change again. State courts follow their own evidence rules too.
The concern behind the proposal is the useful part. People worried about error, bias and machine output reaching a jury without a qualified person to explain or defend it. Your internal chronology is different from trial evidence, and the same discipline is worth having before anything gets that far.
Ethics guidance points the same way
Mata v. Avianca said AI assistance is not improper in itself and placed a gatekeeping role on attorneys. Arizona’s guidance cites a West Virginia opinion saying AI use “requires human input, human oversight, and when necessary human intervention to correct mistakes.”
So does the clinical evidence
Authors of a JAMA Internal Medicine discharge summary study concluded that AI narratives with human review may be a viable option, while reporting more errors in the AI versions. Their conclusion describes assistance, with a clinician in the workflow.
Assisted or dependent: the practical test
Ask two questions about any AI step in your practice. If the output were wrong, who would notice before it left the building? Could that person explain the reasoning to a judge or a client without the software in the room?
If the answer to either is no one, the firm is depending on AI. If a qualified person can answer both, the firm is using AI as an assistant. At Breen, that person is a nurse.
Frequently asked questions
Is proposed Federal Rule of Evidence 707 in effect?
No. In May 2026 the Advisory Committee on Evidence Rules reported it did not recommend action on the proposal at that time and planned further study of a revised draft. Check the committee’s current materials for the latest status.
What is the difference between AI-assisted and AI-dependent work?
In AI-assisted work, a qualified person reviews the output and can defend it without the software. In AI-dependent work, no one would catch a mistake before it leaves the firm.
Sources
- Federal Rule of Evidence 702, Legal Information Institute, Cornell Law School
- Advisory Committee on Evidence Rules, Report to the Standing Committee (May 17, 2026)
- Mata v. Avianca, Inc., No. 22-cv-1461, Opinion and Order on Sanctions (S.D.N.Y. June 22, 2023)
- Supreme Court of Arizona, Steering Committee on AI and the Courts, Generative AI: Ethical Best Practices for Lawyers and Judges (Nov. 14, 2024)
- Williams et al., Physician- and Large Language Model-Generated Hospital Discharge Summaries, JAMA Internal Medicine (2025)
General information for legal professionals. This is not legal or medical advice.