Menu

Terms of use for the free AI check

Version 1.0, effective from 3 October 2026. By clicking "Check for free", you agree to these terms. You can save or print them.

The check is provided by Peak Promotion, the sole-trader business of Joppe Alting Siberg, Address: Generaal Kraijenhoffstraat 36, 1411 BE Naarden, Chamber of Commerce 42128588, info@peakpromotion.nl.

1. What the check is

  1. The check automatically assesses how well AI services such as ChatGPT and Google can read a website. You get an indicator for each check.
  2. The report is generated automatically. I do not review it. It is a snapshot showing the site at the time of the check.
  3. The report is an indication, not advice or a guarantee. The indicators measure technical conditions. They do not predict whether a business will be mentioned in ChatGPT, Google or another service.
  4. A successful check also includes a short AI-generated explanation, unless the site refuses AI use (see 5.4). That explanation may contain errors. It is labelled "Explanation generated with AI" and marked in a machine-readable way (data-ai-generated="true" in the page and ai_generated: true in the data). If the AI service is temporarily unavailable, you receive a fixed explanation I wrote myself. It has no AI label.

2. Free

  1. The check costs nothing. You do not need to provide an email address, create an account or consent to advertising to see the result.
  2. I do not use check data for advertising. If you send a message through the check, you only receive an answer to your question. If it is busy and you leave your email address, I only use it to send the report.

3. Fair use

  1. You can check one new website per internet connection every 30 days. An already checked site whose report is still stored (see 6.1) does not count. An unsuccessful check (see 6.1) does not count either. You can view the same site again unless its owner has objected (see 7.3) or it is temporarily busy (see 3.2). To check another site, send me a message.
  2. Limits per network and per day also apply. If a limit is reached, you can leave your email address.
  3. You may not use the check automatically or in large volumes. You may not bypass its limits or security, for example by repeatedly changing your IP address.
  4. Preferably check your own site or one for which you have the owner's permission. Only check someone else's site for a reasonable purpose.
  5. If you do not follow these rules, I may refuse you access to the check.

4. Reports about other people's sites

  1. A report about another person's site is an automatic snapshot, not a judgement of that business.
  2. Do not use a report to damage another business's reputation. Do not share it as an established fact about that business.

5. What the check fetches

  1. The check only fetches public pages without logging in. At most, it makes these requests in this order:

    • /robots.txt, over https, and over http only if a connection over https fails;
    • /.well-known/tdmrep.json (only to see whether the site has a TDM reservation);
    • the homepage, over https, and over http only if a connection over https fails;
    • whether there is a /sitemap.xml (only whether it exists);
    • whether there is a /llms.txt (only whether it exists).

    If the site redirects to another address, the check follows at most 3 redirects. If a redirect leads to a different domain, the check stops there: you receive "This site redirects to another domain. Enter that domain to check it." The check does not follow links on the page.

  2. The check always identifies itself with this name (User-Agent):

    PeakPromotionCheck/1.0 (+https://peakpromotion.nl/ai-check#site-eigenaren)

  3. The check respects robots.txt. If it blocks the name PeakPromotionCheck on /, or all bots (*) on / without a specific rule for PeakPromotionCheck , the check fetches nothing afterwards. The report only says: "This site refuses automated analysis. I therefore do not check it." If robots.txt cannot be reached (a server error on the site's side, or no connection, including over http), or does not respond in time, the check also fetches nothing else. The report only says: "The site's robots.txt could not be reached. I therefore fetch nothing now." If robots.txt does not exist, no rules apply. If robots.txt only blocks a path, for example /sitemap.xml, the check does not fetch that file; that check is marked "not measured".

  4. The check sends no text to an AI service if the site refuses this in any of these ways:

    • Content-Signal: ai-input=no in robots.txt;
    • a TDM reservation (tdm-reservation: 1 as a header or meta tag, or in /.well-known/tdmrep.json; if robots.txt blocks that file, the check does not fetch it and only looks at the header and meta tag);
    • noai in the X-Robots-Tag or in the robots meta tag.

    You then only receive fixed checks, without an explanation.

  5. The check never bypasses security. If the site returns a refusal (401 or 403) or a bot challenge, it counts as "blocked". The check only recognises a login screen, CAPTCHA or paywall if the site returns such a refusal or challenge; it does not inspect the page content. It does not try again. For "too many requests" (429) or "unavailable" (503), it stops immediately, including if this happens at robots.txt .

  6. The homepage's visible text only stays in memory. It goes to Cloudflare Workers AI once for the explanation. Cloudflare processes the text to generate the answer and does not use the content to train models without explicit consent. I do not keep the text itself. I remove email addresses, phone numbers and bank account numbers before sending it. The report does not show literal passages from the site.

6. How long a report is kept

  1. For each domain, I keep the check results and AI explanation for 30 days. This also applies if the site refuses AI use (5.4) and you only receive fixed checks: that is also a successful check. I keep an unsuccessful check for 1 hour. Unsuccessful means: the site refuses automated analysis (5.3), robots.txt cannot be reached (5.3), the site is blocked, the check stops (5.5), or the site redirects to another domain (5.1). If a domain is on the blocklist (7.3), I keep nothing. Another check of the same domain during that period shows the same report. After deletion, a report may remain in an automatic backup for up to 30 days.
  2. A report is only visible to the requester and me. If you request a check through the form, I run it myself and email you the report. That email remains in my mailbox; its retention period is in my privacy notice (part 1). There is no link to share a report. I do not publish reports.
  3. How I handle personal data, such as your IP address, is explained in my privacy notice.

7. For owners of a checked site

  1. If you do not want the check to fetch your site, put this in your robots.txt:

    User-agent: PeakPromotionCheck
Disallow: /

    The check then only fetches your robots.txt and stops afterwards.

  2. Want a report about your site removed, or object to the check? Send me a message, or email info@peakpromotion.nl.

  3. I aim to reply within 2 working days and handle it within 1 month at the latest. I remove the report and add your domain to a blocklist so it is not checked again. Anyone entering your site then only receives "This site is on my blocklist. I therefore do not check it." Your domain remains on that list for as long as your objection applies. To have it removed, email me.

  4. I never name a checked business in my advertising or on social media without its permission. I do not use the check or its data to approach businesses or for advertising.

8. Changes or discontinuation

You cannot derive rights from the check. I may change or discontinue it.

9. Liability

  1. If you use the check for your business or profession, I am not liable for damage caused by relying on a report, except in cases of intent or deliberate recklessness on my part.
  2. If you are a consumer, paragraph 1 does not apply. I am then liable under the law.

10. Applicable law

Dutch law applies to these terms.