Person crawling under a bed in a dim room, reaching toward the foreground with a flashlight nearby.

Can ChatGPT read your website (and do you want it to)?

Yes, ChatGPT can most likely read your website right now. The quick technical answer is if your site is built like most sites, the machines can read every word of it. If your pages get assembled by JavaScript in the browser, they probably can’t. If your robots.txt file blocks GPTBot, then that crawler is blocked. That’s about it, and you can verify it yourself in two minutes without paying anyone.

But the question I’ve started to ask is, “Do you want ChatGPT reading your website?”

Every marketing newsletter in your inbox this year is teaching you to please the machines. The industry calls the work AEO, GEO, or LLMO, depending on who’s selling it. Hell, even while I was writing this article, I got two LinkedIn notifications for AI Search Specialist job postings. We can’t escape the reality of AI in our world right now (more on that later).

I run a brand agency, I built a free tool that grades AI readability, and I make money when you care about this subject. I’m telling you that up front because everyone selling AI-readiness services has the same conflict of interest.​

Since I run a business, allow me to do my due diligence and walk you through a quick check if your site is visible to ChatGPT before I get into my personal hot take on the topic.

The two-minute check

Open any page on your site and view the page source. Search for a sentence you know is on that page. If you find it in the raw HTML, machines can read you. If you find a wall of code and none of your actual words, your site is invisible to most AI crawlers, because they don’t run JavaScript, and they only read what your server sends.

Then type yoursite.com/robots.txt into your browser. If you see GPTBot or CCBot next to the word Disallow, then someone shut the door. The block might have come from you, from a security plugin, or from a developer in 2023 who read a scary article and panicked.​

Those two checks are more or less the entire audit that some agencies are charging four figures for. If you want the deeper version, I built a free grader that runs them through structural checks (schema, answer placement, crawlability) and scores the results. There’s no email wall, because I hate email walls as much as you do. That grader covers the AI side. In terms of how the same page looks to Google, the free SEO audit runs the ranking checks and scores them the same way.

What actually matters

What matters is the HTML that the crawlers can read. Crawlers look for answers that sit near the top of your pages instead of being buried at the bottom. The llms.txt file that half the industry is excited about belongs at the bottom of the list of things to implement. If an agency leads its pitch with llms.txt, ask them why the fundamentals came second. Keeping those fundamentals in place as your site grows is the ongoing part of our SEO services, and it’s the same work whether the reader is Google or a chatbot.

Why would you let the machines in

People now ask ChatGPT the questions they used to ask a friend: who do you know does brand strategy in Montana? or what a two-day video shoot should cost, and whether the company they’re about to call is legitimate. When an AI answers one of those questions with your name, you have just received a referral from a source your buyer already trusts, and you weren’t even there.​

Google’s AI Overviews now sit atop more than half of searches, and AI referral traffic has been climbing every month since 2025. The people who arrive from an AI answer are pre-sold, because the machine has already explained why you fit. If you sell a service, your thinking is your marketing, and a machine that quotes your thinking is doing that marketing for you around the clock. That’s the honest case for staying open, and it’s the reason my own site is wide open.

Why the other camp is not wrong

Here’s the part the AEO crowd leaves out of the webinar. When AI reads your site, your work becomes the answer, and the answer gets served without you. Google can resolve a searcher’s question using your content on Google’s page, while you get a citation in small text. You don’t get the visit, and you certainly don’t get the sale.

If your content is your product, that trade drains the exact thing you sell. Photographers, writers, publishers, and course creators who let the machines read their catalog are letting the machines serve that catalog without them. And plenty of owners hold an objection that has nothing to do with economics: they built something with their hands, and they don’t want it fed into the machine. I won’t argue anyone out of that, because it’s a coherent position, and it’s held with more conviction than most of the marketing advice published this year.

What I did with my own site

Our robots.txt leaves every AI crawler alone, and that’s a decision I made deliberately. Nobody hires a brand agency from inside a chatbot window, but they do get a name, they might look us up, they might read something like this, and they might book a call. An AI citation costs me nothing and hands me a free introduction to a potential buyer.​

But that’s my business model, and yours may argue for the opposite. The sites that worry me are the defaults on both ends: businesses that are accidentally invisible to AI while their competitors get profiled, and creators who are feeding their life’s work to the models without ever having chosen to. Those are the same mistakes, wearing different clothes.

If your answer is no

Add these lines to robots.txt: User-agent: GPTBot and Disallow: /, then repeat the pair for other bots. Two caveats, though, robots.txt is a request rather than a lock, and it recalls nothing the models have already collected. Google’s AI answers draw on the same index as its regular results, so removing your content from AI answers entirely would mean blocking Google itself and being de-indexed from Google’s Search. It’s a double-edged sword.

Run the free grader, see what the machines currently see, and then make the call the way you’d make any other ownership decision.

My hot take

The world changed, and we’re forced to participate in it. Nobody voted on whether AI belonged in every creative tool we open, and nobody checked with the people who built careers on craft skills before generative fill appeared next to the brush in Photoshop. Even the mechanics of AI happened without including the creators. One example that frustrates me the most is a design request that used to be handled by your computer (using your PC’s CPU and GPU) now travels to a data center to get processed, and it uses an ungodly amount of resources. What a waste. The resources before were just my local PC’s hardware. I was okay with that.

Here’s what I believe happens next. When machines can produce infinite results, those results eventually lose their price and value, and the market starts paying for the one thing machines can’t manufacture: proof that a person was here. That proof looks like the photograph someone took while standing in a river at six in the morning. It’s the proof that a person was present and created something from nothing.

So let the machines handle the boring parts if you want. I would just hold onto the work that only you could have made. That’s the work people will want, and it’s the work the free market will pay a premium for. I’m making that bet with my own company, and I think it’s the right bet for yours.

Common questions

Can ChatGPT read my website?
Yes, if your content is served as real HTML, your robots.txt doesn’t block GPTBot, and the crawler can reach your pages. Sites built entirely in JavaScript are the main exception, because most AI crawlers don’t execute scripts. The two-minute check: search your page source for your own copy, then look at robots.txt.
How do I check if AI can read my site?
View a page’s source and search for a sentence from the page. If it’s in the raw HTML, machines can read it. Then check yoursite.com/robots.txt for lines blocking GPTBot, ClaudeBot, or CCBot. Our free AI readiness grader runs these checks and scores the result.
Do AI crawlers read JavaScript?
Mostly no. GPTBot and most AI crawlers read the HTML your server sends and skip or mishandle JavaScript rendering. If your pages assemble in the browser, assume AI sees a fraction of them. Server-rendered pages, which include most WordPress sites, are readable by default.
Should I block GPTBot?
It depends on what your content does for you. If content is your product, blocking protects it from becoming free answer material. If you sell services, AI citations work like referrals and blocking costs you introductions. Decide from your business model, not from the general mood about AI.
Does blocking AI crawlers hurt my SEO?
Blocking GPTBot, ClaudeBot, and CCBot has no effect on Google rankings. The complication is Google itself: its AI Overviews draw on the same index as regular search, so you can’t exit AI answers while staying in Google results. Blocking Googlebot would remove you from both.
What is llms.txt?
A proposed file that gives AI systems a plain-text guide to your site’s key content. It’s cheap to add and can’t hurt, but it’s a minor signal next to readable HTML, clear answers, and schema markup. Do the fundamentals first.
How do I get my business mentioned by ChatGPT?
Be the clearest answer to the questions your buyers ask. That means pages that answer one question directly, schema markup, consistent business information across the web, and mentions on sites AI systems trust. It’s the same work that earns featured snippets, which is why we treat SEO and AI visibility as one discipline.
lets canoe logo reversed

We checked. Nobody remembers boring marketing.

Welcome to a brand agency that leads with strategy.

Let’s start with a comprehensive brand strategy audit.

brian and anna murnion founders lets canoe brand agency
Let’s Canoe founders Anna and Brian Murnion.

© 2026 CANOE, LLC dba Let’s Canoe. All rights reserved.