vernac / architecture, expressed.
Browse documentation

AI content disclosure review for vernac.org

Status: brief assessment completed on 2026-10-15; user chose no additional AI-content labels for the current described publication workflow.

The official Article 50 text and Commission transparency FAQ were checked, together with the Bundesnetzagentur’s transparency overview. The assessment distinguished provider marking duties from publisher/deployer disclosure, public-interest text with the human-review/editorial-responsibility exception, deepfake media and source code. It found no blanket per-file disclosure requirement for the described human-reviewed technical documentation, code and schematic diagrams. This was not a complete audit of every published asset or a binding legal opinion.

Reassess if publication becomes unreviewed/automated, realistic synthetic media is added, or Vernac distributes an AI system. No website banner or voluntary notice is introduced by the user’s decision. The scope below records the original review brief.

Question to resolve

Vernac.org and the public Vernac project contain AI-assisted or AI-generated prose, illustrations and code, with human review and editing. Determine whether German or EU requirements impose disclosure, marking or other transparency duties on this specific publication and its operator, and what concrete action is required.

Investigation scope, when this work starts

Expected deliverable

A concise, source-backed assessment for the actual Vernac setup and, where indicated, a reviewable patch for the appropriate English website/documentation notices or markings. Avoid blanket statements that misrepresent how individual content was created. This work does not itself change Vernac’s project or extension licensing.

The scheduling dependency is recorded in the IDE/debugger roadmap.