GitHub Copilot was linked to more open-source code—and slower coordination

Projects saw more contributions and timely code merges, but developers also spent more time discussing and coordinating changes.

Editorial illustration — not from the study.

Researchers combined GitHub’s proprietary Copilot usage data with public data from open-source projects to examine how the AI coding tool relates to collaborative software development. Copilot use was associated with increases in project code contributions, developer participation and individual contributions.

The tool was also associated with more code discussions and an 8% increase in coordination time. Peripheral developers saw smaller gains in project-level contributions and larger increases in coordination time than core developers, although projects overall experienced a net increase in timely code merges.

More code, more coordination

Copilot use was associated with a 5.9% increase in project-level code contributions, a 3.4% increase in developer coding participation and a 2.1% increase in individual code contributions. At the same time, coordination time increased by 8%, and projects had more code discussions. Peripheral developers had relatively smaller increases in project-level contributions and larger increases in coordination time than core developers. Despite this tradeoff, Copilot use was associated with a net increase in timely code merges at the project level.

The tradeoff for open source

The findings point to two effects of AI coding tools in volunteer-run software projects. They can broaden participation and increase the amount of code contributed, while also adding work to the process of reviewing, discussing and coordinating changes. The larger coordination burden for peripheral developers suggests that the effects of these tools are not shared evenly across open-source communities.

What the data can show

The study combines GitHub’s proprietary Copilot usage data with public data on open-source projects. The abstract does not specify the number of projects or developers, the study period or the detailed research design, so it does not establish how broadly the results apply beyond the projects examined. Because the abstract describes relationships in usage and project data rather than a randomized test, the findings should not be read as proving that Copilot alone caused every reported change.

A review finds human-like AI still faces major technical barriers

A systematic literature review catalogs unresolved technical challenges in developing artificial general intelligence, including limited common sense, unreliable responses and difficulty handling unfamiliar situations. It points to hybrid systems, embodied approaches and stronger reasoning as possible research directions, while emphasizing the need for better safety and verification.

20 hours ago·4 views

AI & Computing

An AI system outlines nearly all objects in huge images of the Earth

A new open-source pipeline uses an existing AI system to segment far more of very large Earth-observation images without training data or manual threshold tuning. Across seven scenes, it increased coverage to 91–98% and joined objects split across image tiles.

1 day ago·39 views

AI & Computing

Preprint reports a complete check of the 14-runner lonely runner problem

A new preprint reports a complete computer-assisted verification of the Lonely Runner Conjecture for 14 runners, with 13 of them moving. The archive includes certificates, source code and an audit script; because it is a preprint, peer review may change the findings.

1 day ago·10 views