AI assistants boost developer productivity by 26.08%, study finds

How the study measured productivity gains

The researchers conducted three randomized controlled trials in real corporate environments. In this methodology, participants are randomly assigned to groups that either receive or do not receive the technology, allowing for a direct comparison between those who used the AI programming assistant and those who worked without it.

The assistant provided intelligent suggestions for completing code snippets, with the aim of evaluating its impact on programmers’ daily work under real-world business conditions.

The combined results from the three experiments indicated a 26.08% increase in the number of completed tasks. However, researchers caution that statistical variations and differences in individual outcomes prevent this percentage from being uniformly applied to all professionals.

Less experienced professionals saw greater gains

Data analysis revealed that less experienced developers had higher adoption rates and more significant productivity increases. One possible explanation is that automated suggestions help overcome steps that still require research and execution time.

The technology can serve as a support tool during the learning phase, but it does not eliminate the need to understand the code and validate its functionality.

For companies, the findings suggest that AI implementation should consider the needs of different team members. Specific training, appropriate task selection, and continuous monitoring of results are key factors for successful adoption.

More completed tasks don’t guarantee quality

The study used the number of completed tasks as the primary productivity metric. While relevant, this indicator alone does not reflect code quality, security, or performance after implementation.

A tool can speed up code writing, but incorrect suggestions still require correction. Therefore, any AI evaluation must also include analysis of errors, rework, review time, and subsequent maintenance.

The results confirm that programming assistants can increase productivity in real professional settings, but they should not be automatically extrapolated to all occupations or any type of AI solution.

How to integrate AI into work effectively

The first step for professionals is to identify tasks where technology can add value and test its performance using clear criteria, such as execution time, code quality, and the volume of corrections needed.

Familiarity with tools should be accompanied by technical knowledge and critical thinking skills. Knowing how to formulate precise instructions, recognize errors, and decide when not to use an AI-generated response are essential competencies in this process.

Training course offers practical AI tools

Professionals interested in deepening their use of AI can enroll in a training program that includes four online practical lessons—three recorded and one live—focusing on around 20 tools, including ChatGPT, Claude, NotebookLM, and Manus.

The content covers autonomous research, document analysis, creation of custom agents, voice cloning, and video production with avatars. The course includes a completion certificate and 30 days of access to Inner AI Pro, a paid platform offering over 50 AI tools. The investment is R$ 37.