81% of Developers Spend More Time on Code Reviews

Harness’s 2026 State of Engineering Excellence report found 81% of developers now spend more time on code reviews after adopting AI tools. The report says many teams are not recording the extra remediation and validation work that follows AI integration.

The study reports that 28% of developers say code review tasks now take about 30% longer on average. Nearly one-third of that activity-an estimated 31% of developer time-is not tracked by teams. Untracked work includes reviewing AI-generated code, fixing bugs introduced or missed by AI, and switching between uncoordinated tools.

The survey highlights a gap between standard productivity metrics and AI-influenced workflows. About 89% of tech leaders still rely on measurement frameworks that do not reflect AI’s impact on individuals or teams, and 94% said metrics omit factors such as technical debt and developer burnout. When asked to name the single biggest AI-related challenge, respondents cited measuring true productivity impact (26%), maintaining code quality with AI (24%) and proving return on investment to leadership (18%).

Harness recommends that organizations account for follow-up tasks created by AI, including code validation and rework, and measure the quality of code produced by AI assistants. The firm suggests treating AI performance as a separate discipline, with metrics for agent accuracy, acceptance rates for AI-generated outputs and the direct costs of AI tools compared with human output.

Nearly half of developers (49%) said they want to take part in defining performance metrics under the new AI workflows. The report also found a perception gap: managers were nearly four times more likely than frontline engineers to report no concerns about how productivity is being measured.

Trevor Stuart, senior vice president and general manager at Harness, described the findings as a change in how developers spend their working day: “Cloud and the internet were infrastructure revolutions layered underneath the developer. AI is reshaping the developer’s job entirely, and the measurement frameworks that the industry has relied on for the past decade weren’t built for this new unit of work.”

The report states that without updated visibility and new metrics, organizations may underestimate downstream costs of AI adoption, including increased remediation, hidden labor and rising developer fatigue.