What we measured
We examined the websites of 282 Swiss municipalities — from the city of Zurich to mountain communes in Val Poschiavo. We measured what is visible to AI assistants and search engines: public, passive signals, not systems and not content behind a login.
Specifically, this includes:
- robots.txt — may the website be read by machines at all, and are there rules for AI crawlers?
- llms.txt — is there a file that tells AI assistants specifically about the municipality?
- Sitemap — is the structure of the website prepared for machines?
- Structured data — do machines understand what a page represents?
- Chatbot fingerprints — does the website show signs of a chatbot?
- Open data — does the municipality publish machine-readable datasets?
The survey was conducted on 1 October 2026 with our own measurement pipeline. It reads only publicly accessible files and pages — no enquiries were made to municipalities and no systems were touched.
The key results
- Only 2 of 282 municipalities have an llms.txt: Davos and Rorschach.
- Only 2 of 282 have any rules for AI crawlers in their robots.txt — and both forbid them: Schwende-Rüte and Thal.
- 0 chatbots were detected. This, however, is the limit of our static detection: a chatbot loaded only after a click leaves no fingerprint in the static HTML.
- 242 websites have a sitemap.
- 7 organisations publish open data.
- The median is 32 out of 100 points.
An example: Allschwil
Allschwil in the canton of Basel-Landschaft, with around 21,000 residents, is a typical mid-sized municipality. Its website scores 30 out of 100 points, placing it 212th of 282 — a completely normal result, not an outlier on the low end.
This means: the technical visibility of most municipal websites is limited for AI assistants. Citizens who ask an AI about opening hours, forms, or the municipal council today may receive no answer from the source — or an outdated one.
What does this mean?
This is not a reproach to the municipalities. The signals we measure are new, and many administrations have other priorities. But public expectations are changing: search engines were yesterday; AI assistants are tomorrow's first point of contact. Municipalities that are easy to find there give citizens better answers — with the same content, simply better prepared.
The good news: most signals can be improved with manageable effort. llms.txt and structured data are technical details with a noticeable effect.
With the AI readiness check you can assess where your municipality stands across seven dimensions — from strategy and data to citizen communication.