Amber Matz Reflects on Trust and Expertise in AI-Assisted Open Source Contributions

Artificial intelligence is changing how contributions arrive in open source projects, but Amber Matz argues that the resulting bottlenecks are not always technical. In a blog post reflecting on a conversation with DDEV maintainer Randy Fay during DrupalCon Chicago, Matz describes a substantial AI-assisted pull request that appeared technically sound, aligned with project goals, and feasible to review. The hesitation arose elsewhere: uncertainty about trust, long-term stewardship, and the maintenance obligations attached to a significant code change.

The post examines how AI-assisted development is altering traditional signals of trust within open source communities. Contributors can now produce sophisticated code changes with the assistance of large language models, potentially expanding participation and accelerating work on longstanding issues. For maintainers, however, evaluating a contribution increasingly involves questions beyond code quality, including whether contributors can support their changes over time and how project knowledge is transferred and preserved.

Drawing on Fay's experience, Matz argues that concerns about AI-assisted contributions ultimately reflect broader questions about expertise and responsibility in open source software. The article notes that AI-generated contributions can be valuable, but that deep familiarity with a project's behaviour and architecture remains difficult to replicate through generated output alone. Matz frames the resulting uncertainty as part of a wider transition in software development, where maintainers and contributors are reassessing established ideas about trust, stewardship, and expertise.