The AI Market–Readiness Gap →
Key finding: GPT and Gemini carry 83.0% of the current platform weighting, yet only 17.4% of 63,858 crawler-accessible EU-27 company websites pass the structural-readiness threshold for both.
EU AI Readiness Observatory
Research for the AI-mediated web.
Independent evidence on how organisations are accessed, understood and cited by AI systems — with transparent methods and permanent sources.
- 3 published studies
- 1 report · 1 data note · 1 threat note
- Updated 16 Sep 2026
Published research
Newest first · permanent URLs
Some crawlers are thieves in disguise →
Key finding: 10,064 of 12,948 AI-labelled requests could not be provider-verified; 4,355 targeted sensitive technical paths.
A high SEO score does not predict AI readiness →
Key finding: Across 78,110 EU-27 company websites, 68.6% of strong-SEO websites remained below the citation-readiness threshold.
About the Observatory
The EU AI Readiness Observatory publishes independent evidence about how organisations are accessed, understood and cited by AI systems.
- Reports provide large-sample analysis and downloadable datasets.
- Data Notes isolate one decision-relevant pattern from the current dataset.
- Threat Notes publish focused field evidence about a specific risk or behaviour.
- Permanent sources preserve methods, corrections and citation-ready data.
Open evidence
Published aggregated results are available under CC BY 4.0 with attribution to VerisAI, EU AI Readiness Observatory, verisai.eu.
Frequently Asked Questions
What is the EU AI Readiness Observatory?
An evidence programme that measures how organisations are accessed, understood and cited by AI systems.
Where can I find all published research?
Every publication is listed in the Published research index and keeps its permanent URL.
Can I reuse the data?
Yes. Published aggregated results may be reused under CC BY 4.0 with attribution.
Is company-level source data published?
No. Public downloads contain anonymised aggregate results and exclude company names, domains, URLs, IP addresses, scan identifiers and raw page content.
Where is your AI visibility at risk?
See the market exposure, platform readiness gap, methodology and anonymised evidence.