Does Your Cookie Banner Block ChatGPT and Perplexity?

The question almost nobody asks

You've set up a GDPR-compliant cookie banner. Good news on the compliance side. But a question almost nobody asks : does that same banner change what ChatGPT, Perplexity, or Google AI Overviews see of your site?

The short answer : a well-built banner, no. A badly-built one, potentially yes — but not for the reason you'd think.

What a cookie banner actually blocks

A compliant banner (GDPR) blocks third-party tracking scripts before consent : Google Analytics, Meta Pixel, ad pixels. That's the rule — without this blocking, data collection starts before the visitor has even made a choice, which is exactly what regulators penalize.

This blocking never touches your page's HTML content : your text, your headings, your images, your schema.org markup stay present in the source code, visible without any cookie being accepted. An AI agent reading your page — whether a crawler or an agent "browsing" your site to answer a question — sees that content exactly as a visitor who clicked "Reject all" would.

The 3 real mistakes that DO break your AI visibility

1. Schema.org injected by script, and blocked along with everything else

Some setups inject their structured data (JSON-LD) via a JavaScript script — common on certain site builders. If that script gets misclassified by the consent platform and blocked as "non-essential," your schema markup disappears for any visitor who hasn't accepted. An AI agent reading the page without executing all the JavaScript will never see that schema.

The fix : schema.org markup should be injected directly into the served HTML, never conditioned on a script classified as "marketing" or "analytics."

2. A consent wall that hides all the content

Some tools display an overlay that visually hides the entire page until the user interacts — and some go as far as delaying the main content from loading behind that wall. This is a common abuse (sometimes deliberate, to force acceptance) that regulators already penalize for other reasons. An AI agent that needs to "click" to continue, the way a real visitor would, may simply stop there if it finds nothing to act on.

The fix : page content should never depend on an interaction with the banner to be present in the DOM — only tracking-script execution should depend on it.

3. Blocking via robots.txt "just to be safe"

Rarer, but seen in the wild : sites that, out of GDPR over-caution, block indexing entirely (via robots.txt or noindex) until a preference-management system is deemed "complete." GDPR never requires this — it's a confusion between consent to tracking and the right to access public content.

How to check your own site

1. Open your site in a private/incognito window, without accepting anything
2. View the page source (Ctrl+U) — not the inspector, the actual raw source
3. Look for your main text and your <script type="application/ld+json"> tags
4. If they're there, in plain sight, without having accepted anything : you're fine. If they're missing : you've found your AI-visibility gap.

That's exactly what Livada Trace checks, for free — the "Can AI read you?" signal tests precisely this, cookies accepted or not.

What Livada Cookies actually does

Livada Cookies blocks third-party tracking scripts before consent — that's its job, and the law requires it. It never touches your page's HTML, nor the native schema.org markup from your theme or SEO plugin. GDPR compliance and AI visibility aren't competing goals — they're two completely different layers of the same site.

Recap

• A compliant cookie banner blocks tracking, never HTML content
• The real risk : schema.org injected via script and misclassified by the consent platform
• Second risk : a consent wall that delays the main content from rendering
• Check it in 30 seconds : view source, without accepting — your content should be right there
• Test it for free with Livada Trace