Halvus research ยท 200-company benchmark
Do Europe's leading websites respect consent and label their AI?
We opened the websites of 200 leading UK and EU companies in a real browser, recorded every request before any cookie choice, and tested their AI chat, AI hiring tools and AI crawler settings.
Results coming soon
Results will be published here sector by sector once the study is released.
Sectors in the study
| Sector | Companies | Status |
|---|---|---|
| FinTech & neo-banking | 35 | Coming soon |
| Tech, marketplaces & consumer apps | 40 | In progress |
| E-commerce & fashion | 55 | Coming soon |
| Travel & airlines | 40 | Coming soon |
| UK household names | 30 | Coming soon |
Run the same checks on your site
- Cookie scanner: trackers firing before consent
- AI chatbot disclosure checker: EU AI Act Art. 50(1)
- AI recruitment checker: AI hiring notices
- AI crawler opt-out checker: robots.txt and TDM
- Full EU AI Act checker
Common questions
- How were sites tested?
- Each home page was loaded in a fresh real browser from London with no stored cookies. We recorded every network request and cookie before touching the cookie banner, then read the legal pages, careers page and robots.txt. Every finding is backed by a recorded request, cookie or page quote.
- What counts as sending data before consent?
- A request to an advertising network carrying visitor identifiers, or an ad cookie being set, before the visitor accepted. Tags that loaded but sent nothing are counted separately, as a warning.
- Why were no AI chatbots flagged?
- We only flag a chat that is proven to be AI, actually opens on screen, and shows no AI notice.
- Is blocking AI crawlers required by law?
- No. EU copyright rules let site owners opt out of AI training in a machine-readable way. We report it as context, and it never affects the Risk Score.
- Can I see my company's result?
- Yes. Run a free scan of your site above and you'll get the same report we used for the benchmark.