AI coding agents generate more code, but not more software
A study finds AI coding agents speed coding but raise pull-request review time 49% and change requests.
Researchers using Jellyfish engineering data found AI coding agents increase code output but slow delivery because human review expands. Average time from pull request to merge rose 49%, the share of pull requests with requested changes nearly doubled, and comments per pull request rose 35%. Firms saw a 14% rise in workers doing reviews, but no significant employment change versus LinkedIn data. By March 2026, 80% of measured firms used AI code review, yet agents produced only 23.3% of review comments and 10.8% of pull requests; 95% of firms had deployed coding agents.
- Average pull-request review time rose 49% after AI coding agents were introduced.
- Pull requests needing changes nearly doubled, and comments per request rose 35%.
- The share of workers doing code review increased 14%, with no clear job losses.
- AI produced 23.3% of review comments and 10.8% of pull requests by March 2026.
- Ninety-five percent of studied firms had adopted AI coding agents.
From the original article
2 min read at arstechnica.com
Study finds coding efficiency gains get "absorbed" by human review "bottleneck."
By Kyle Orland · arstechnica.com
A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.