Insights / AI in Personal Injury Litigation
How AI Is Changing Medical Record Review for Personal Injury Law Firms
By Breen LNC.
Short answer: AI is changing the first pass of medical record review. It can read, sort and draft from a large record quickly, which matters because records keep growing. One published comparison found more errors in AI hospital narratives than in physician ones, and other studies report omissions, so the draft still needs a clinician before an attorney relies on it.
Why the record became the bottleneck
Medical records have grown much faster than anyone’s ability to read them. One health system study counted the notes available to physicians at emergency visits and found a median of 5 notes in 2006 and 359 in 2022, going from 1,735 words to 58,662. At the University of Pennsylvania Health System, researchers found that 50.1% of the words in more than 104 million notes were exact duplicates of earlier text about the same patient.
Neither study measured a litigation file, and a real case will differ. The direction is still clear. Somebody on your team has to find the few lines that matter inside a record that keeps getting longer and more repetitive.
What AI does well
Sorting, deduplicating, drafting a timeline and pointing to a page are the kinds of tasks software can do far faster than a person. Adoption reflects that. The ABA’s 2024 survey, as reported by the ABA Journal, found 30% of responding lawyers using AI, with concerns about hallucinations cited by three quarters of respondents.
Where the research urges caution
The closest published evidence comes from hospitals, not law firms. In a JAMA Internal Medicine study of 100 discharge encounters, AI narratives were more concise but less comprehensive than physician narratives, with 2.91 errors per summary against 1.82. A 2026 pilot on one inpatient unit saw physicians note omissions in 25 of 100 rated summaries and hallucinations in 2. Omission was the larger problem.
For an injury case that detail matters. A fabricated line can often be caught by checking the page. A missing one gives you nothing to check.
We did not find a published study that measures error rates in litigation chronologies, so treat these numbers as a warning sign and not as a benchmark.
What this means for your firm
Use AI for speed on the first read. Then have a nurse decide what each entry means and whether anything is missing. That is how Breen works: the speed of AI, the judgment of a nurse.
Frequently asked questions
Can AI review medical records on its own?
It can draft and organize, but a hospital study comparing AI and physician discharge narratives reported more errors in the AI versions, and its authors point to human review. An attorney should have a clinician check the output before relying on it.
What kind of AI error is most common in medical summaries?
In a 2026 pilot of AI hospital course summaries, physicians reported omissions in 25 of 100 rated summaries, inaccuracies in 20 and hallucinations in 2. That was one hospital unit, so it is not a general rate.
Sources
- Patterson et al., Call me Dr Ishmael: trends in electronic health record notes available at emergency department visits and admissions, JAMIA Open (2024)
- Steinkamp et al., Prevalence and Sources of Duplicate Information in the Electronic Medical Record, JAMA Network Open (2022)
- ABA Journal, ABA tech report finds that AI adoption is growing but some are hesitant (Mar. 7, 2025)
- Williams et al., Physician- and Large Language Model-Generated Hospital Discharge Summaries, JAMA Internal Medicine (2025)
- Grolleau et al., Physician-Reported Safety Outcomes of AI-Generated Hospital Course Summaries, JAMA Network Open (2026)
General information for legal professionals. This is not legal or medical advice.