Does AI Assistance Enhance or Erode Expertise? Evidence from a Three-Month Field Experiment in Patent Drafting
NBER Working Paper 35720, September 2026
Working paper · not peer reviewed
What it adds
Tests assisted drafting and later unassisted patent review in a preregistered experiment with practicing lawyers, using blinded expert grading to distinguish work quality from retained expertise.
Read with this limitation
Only 91 of 133 participants completed the protocol. The selected firms had Google relationships; Google funded the study. No baseline skills assessment was possible, and unassisted-task compliance was imperfectly observable. Benchmark-task results and a three-month horizon do not establish long-term client outcomes.
- Jurisdiction tags
- United States
- Language
- English
Research artifacts
Resources linked by the authors.
Follow the source
Selected statements located in the linked primary sources. These results have not been independently reproduced here.
The experiment randomized 133 patent lawyers from eleven US firms; 91 completed the full protocol.
1 IntroductionAfter three months, the reported gains on unassisted patent redlining were concentrated among senior lawyers; juniors showed no average gain.
6 Discussion
Connected research
Related question
AI-Powered Lawyering: AI Reasoning Models, Retrieval Augmented Generation, and the Future of Legal Practice
Both examine AI-assisted legal work. The patent-drafting experiment adds sustained exposure and an unassisted assessment of practicing lawyers. The connection is thematic, not a citation claim.
Entry status, source edition & rights
- Index entry
- Source-linked summary, prepared with agent assistance. Independent scholarly review is pending.
- Archive admission
- This bibliographic entry does not confer admission to the versioned archive.
- Edition inspected
- 10.3386/w35720
- Reading coverage
- September 2026 NBER working-paper PDF: author information, introduction, study design, results and discussion inspected, including PDF pages 2–4 and 38–40. DOI and month verified against publisher-deposited metadata. The paper expressly states that it has not been peer reviewed. No experiments or artifacts executed.
- Rights
- External source and artifact links only. No full text or research data is mirrored; original-host terms apply. An artifact link does not establish access or reproduction.
- Entry updated
- 2026-10-15
Source-linked summary · independent assessment pending.