[Research: AI Code Review Can Reduce Review Time, but Quality Risks Cannot Be Ignored] DeepTech TechFlow reports, on July 27, according to a research summary published by Rohan Paul, a large-scale study involving 207 GitHub projects and 1.02 million pull requests revealed that introducing AI agents for code review can shorten review time by 2.5 to 4.5 days/KLOC. However, this comes at the cost of reduced review quality—78% to 94% of PRs reviewed with AI exhibited 'review smells,' compared to 69% to 76% in purely human reviews. The study pointed out that repeatedly assigning the same AI reviewer identity is the primary reason for the decline in review diversity. Notably, projects that heavily adopted LLM reviews in their early stages did not achieve significant efficiency gains.