GitKraken Research: AI Boosts Developer Output 25%—But Also Increases Code Churn

SCOTTSDALE, Ariz. - March 26, 2026 - PRLog -- GitKraken in partnership with GitClear today released new research, "The AI Multiplier Effect," analyzing how AI coding tools are changing day-to-day engineering work.

The data shows a clear pattern: AI is increasing developer output—but not without tradeoffs.

During a three month period, across 2,172 developer-weeks using tools like GitHub Copilot, Cursor, and Claude Code, GitKraken and GitClear found that AI acts as a multiplier on existing developer behavior. It helps teams move faster, but it also introduces new challenges that aren't captured by traditional productivity metrics.

What the data shows

• Developers see 25% productivity gains when compared to their own historical performance
• Heavy AI users generate 4–14x more activity than low or non-users
• Test code generation increases 4x, improving coverage in many cases
• Code churn increases up to 9x, pointing to higher rework and potential technical debt
• Review effort scales more efficiently, with output increasing faster than review time
• Anecdotally, developers who have leaned into increasingly autonomous agentic workflows in the past six months have observed exponential productivity gains

Taken together, the results suggest that AI is not magically creating the mythical "10x engineers"—but it is making engineers who adopt and adapt much faster, more productive, and more iterative.

"Teams are seeing greater output, but they don't have an effective way to accurately quantify the gains, or evaluate whether that output is actually better," said Jeremy Castile, Vice President of Developer Research. "The challenge now isn't adopting AI—it's understanding its impact so you can grow its impact."

Output is up, but so is rework

One of the clearest signals in the data is the rise in code churn. As developers generate more code with AI, they also revise and replace more of it.

That creates a new kind of cost: not in writing code, but in reviewing and maintaining a growing code base over time.

The report also finds that differences between developers still matter. Much of the variation in output exists regardless of AI usage, reinforcing that AI amplifies existing skill levels and throughput, rather than equalizing them.

Why existing metrics fall short

Most engineering teams still rely on metrics designed before AI-assisted development was common. Those metrics track activity, but they don't fully capture quality, rework, or long-term maintainability.

As a result, teams can see higher output without a clear understanding of whether they're improving overall performance—or just shifting where the work happens.

A shift toward outcome-based measurement

The report suggests a shift in how teams evaluate productivity:

• Measure quality alongside speed
• Track churn and rework as first-class signals
• Connect engineering output to business outcomes like delivery time and reliability

GitKraken Insights builds on this approach by combining AI usage data, code quality signals, DORA metrics, and developer experience into a single view.

The goal is simple: help teams understand not just how fast they're moving, but whether they're moving in the right direction.

About the research
The analysis is based on a three month analysis of 2,172 developer-weeks of activity across multiple AI coding tools and real-world development environments. The study compares developers against their own prior performance to isolate the impact of AI.

About GitKraken
GitKraken provides developer tools used by more than 40 million developers worldwide. Its products—including GitKraken Desktop, GitLens, and GitKraken Insights—help teams collaborate, improve code quality, and ship software more efficiently.

Read the full report:
https://www.gitkraken.com/reports/ai-multiplier-effect