Download now

Executive Summary

AI boosts software development productivity up to 40% for simple tasks on greenfield projects, but just 0-10% on complex tasks for brownfield applications, as LLMs lack context of intricate dependencies in legacy code.

In this study, Stanford AI Lab researcher Yegor Denisov-Blanch analyzes millions of GitHub commits from 100,000+ engineers to measure AI’s real-world impact on development. The findings show that while AI increases code output, it also increases rework, delivering diminishing returns as codebases grow in size and complexity.

Download the study to understand where AI is creating meaningful productivity gains, and where software architecture becomes a limiting factor.

Yegor Denisov-BlanchStanford AI Lab
Yegor Denisov-Blanch

Research Scientist

Yegor Denisov-Blanch studies how artificial intelligence is changing software engineering. His research focuses on measuring real-world engineering productivity, AI adoption, code quality, and organizational outcomes across large populations of repositories and teams. He designs empirical methods and metrics that move beyond simple proxies to accurately quantify software output, rework, and AI-assisted development at scale.