Drupal AI Experiments Test a Path to Community Projects

Personal AI experiments can become candidates for broader Drupal use after they are published and used or improved by others, according to Drupal founder Dries Buytaert. In From personal AI experiments to shared tools, published 28 July 2026, Dries recaps an hour-long presentation to the Drupal AI Learners Club. The club was started by Angie Byron and is co-organised with Amber Himes Matz as a forum for demonstrations and discussion of AI use in Drupal development and site building.

The first demonstration covered Drupal Digests, which monitors issue queues for Drupal Core, Drupal CMS, Drupal Canvas, and the Drupal AI initiative and publishes AI-written summaries of noteworthy changes. Dries also showed how that pipeline can analyse Drupal Core issues and code changes to generate Rector rules for automating parts of an API upgrade. A third experiment uses an API Catalog and an OpenAPI description to help software agents discover and call the search service on his website.

Together, the experiments address three recurring needs: following changes across active Drupal projects, applying API migrations, and making website capabilities discoverable to agents. Dries says he publishes tools that appear useful beyond his own workflow and then watches whether people use them or contribute improvements. He writes that this public use can show whether an experiment may eventually become a community project, and the post links to the session recording for demonstrations and participant questions.