From Human-Centric to Agentic Code Review: The Impact of Different Generations of Generative AI Technology on Review Quality

This large-scale empirical study examines 1.02 million pull requests across 207 open-source GitHub projects spanning three code review eras, tracing how successive generations of generative AI technology have reshaped review practices.

Key findings:

• The study identifies three distinct AI adoption patterns in how projects integrate generative AI into code review
• Agent-involved collaboration patterns — especially reviews initiated by AI agents or involving multiple AI agents — are associated with faster review decisions under certain adoption strategies
• These efficiency gains do not correspond to improved review quality
• Once AI systems participate in the review process, human-AI collaboration patterns become the strongest explanatory factor for review efficiency

The study offers evidence-based recommendations for designing AI-supported code review systems that improve efficiency while safeguarding review quality.

Authors: Suzhen Zhong, Shayan Noei, Bram Adams, Ying Zou
arXiv: 2607.13196