“Reviewers are overloaded with the increasing volume of code from AI-assisted code development, and have limited time to identify subtle bugs, security issues, or consistently enforce best practices.”
Quote · uber.com
From “uReview: Scalable, Trustworthy GenAI for Code Review at Uber”, found 2026-10-14.
uReview reviews over 90% of the roughly 65,000 diffs Uber lands each week. Its main lesson is that precision matters more than volume: once reviewers see too many false positives, they ignore the bot.