AI Tools May Slow Experienced Developers Instead Of Boosting Productivity, Study Finds Tasks Take 19% Longer With Assistance: Report

An experiment shows that AI tools, widely touted as productivity boosters, may actually slow down experienced software developers instead of speeding up their work.

AI Tools Increase Task Completion

In July 2025, in the study conducted by nonprofit research group Model Evaluation and Threat Research (METR), 16 software developers with an average of five years' experience completed 246 tasks, half with AI assistance and half without.

Developers used tools such as Cursor Pro and Claude 3.5/3.7 Sonnet.

Participants had expected AI to reduce task time by 24%, but instead, tasks completed with AI took 19% longer than those done without it.

"While I like to believe that my productivity didn't suffer while using AI for my tasks, it's not unlikely that it might not have helped me as much as I anticipated or maybe even hampered my efforts," participant Philipp Burckhardt wrote in a blog post.

Developers Spend Extra Time Debugging And Adapting AI Outputs

Researchers attributed the slowdown to AI's lack of project context, reported Fortune.

Developers often had to adapt and debug AI-generated outputs, write prompts, or wait for results.

"AI outputs that are generally useful to them… developers have to spend a lot of time cleaning up the resulting code to make it actually fit for the project," said study author Nate Rush.

See Also: Trump’s AI Order Under Fire— Amy Klobuchar Says It’s ‘Likely Illegal’ While Bernie Sanders Calls It ‘Extremely Dangerous’

AI Productivity Gains Raise Concerns For Workers

Last year, studies showed that AI tools boosted productivity in some cases but also created challenges for employees.

Anthropic's study found Claude Code helped workers complete more tasks and delegate up to 20% of repetitive work, but collaboration declined, skills eroded, and some feared AI could make them irrelevant.

A separate study of 25,000 Danish employees using AI chatbots between 2023 and 2024 found no significant changes in wages or hours worked, suggesting generative AI like ChatGPT had yet to deliver broad productivity gains and sometimes added extra work.

Sen. Bernie Sanders (I-VT) warned that AI should benefit workers, not just billionaires, citing predictions that it could eliminate half of entry-level white-collar jobs within five years.

Read Next:

• Trump Administration Locks In Zero Tariffs On UK Pharmaceuticals As NHS To Pay More For US Drugs

Disclaimer: This content was partially produced with the help of AI tools and was reviewed and published by Benzinga editors.

Photo courtesy: Shutterstock