Outrings
AI visibility and answer engines

Does blocking GPTBot hurt my traffic?

Blocking GPTBot alone costs you almost nothing. The damage comes from the other rules people add at the same time.

3 min read
Short answer

Blocking GPTBot by itself has no measurable effect on traffic, because GPTBot only collects training data. The traffic loss people attribute to it comes from also blocking OAI-SearchBot and ChatGPT-User, which are what make you visible and reachable inside ChatGPT.

What GPTBot actually does

GPTBot crawls pages to build training corpora for future models. Its work does not affect what ChatGPT can retrieve today, does not affect whether you are cited, and does not affect search ranking anywhere. Block it and the only consequence is that your content is excluded from future training runs — which may be exactly what you want.

What people block by accident

The traffic cost comes from the neighbouring rules. Two crawlers do the work that produces visits:

  • OAI-SearchBot builds the search index ChatGPT retrieves from. Blocked, you cannot appear in ChatGPT search results at all.
  • ChatGPT-User fetches a page live when a user asks about it. Blocked, someone who pastes your URL gets an error — the most engaged possible visitor, turned away.

These get blocked in two ways. Either a copied snippet lists all three, or an existing User-agent: * / Disallow: / already covered them and the specific GPTBot rule was added on top of an already-closed door.

Check what you have

  1. Open yoursite.com/robots.txt.
  2. Look for any group containing Disallow: /, especially under User-agent: *.
  3. Confirm there are explicit Allow: / groups for OAI-SearchBot and ChatGPT-User if you want to remain visible.
  4. Test by pasting one of your URLs into ChatGPT and asking what the page says. If it cannot read it, ChatGPT-User is blocked.

The configuration you probably want

# Opt out of training
User-agent: GPTBot
Disallow: /

# Stay visible and reachable
User-agent: OAI-SearchBot
Allow: /

User-agent: ChatGPT-User
Allow: /
One caveat about measurement. Referrals from AI assistants are systematically undercounted: many arrive with no referrer header at all, so they land in "direct" traffic. If you are judging whether AI visibility is worth anything by looking at a referrer report, you are probably looking at an understatement.

Deciding honestly

The question is not really about traffic. It is whether you object to your content being used for training. If you do, block GPTBot and keep the retrieval crawlers open — you lose nothing visible. If you do not object, allowing it costs you nothing either. What you should not do is block all three by accident and then conclude that AI visibility does not work.

What our audit reports about this

Every item below is measured directly, not inferred. Run it against your own site and the result names the exact rule or header responsible.

  • GPTBot, OAI-SearchBot and ChatGPT-User reported separately, each with the rule that governs it.
  • Whether a catch-all rule is blocking crawlers you believe you allowed.
  • Whether the robots.txt being served matches what you published.
  • Whether anything else — a noindex tag or header — would remove you regardless of robots.txt.

For agents and scripts, the same measurement is at /api/v1/ai?url=yoursite.com — see the API documentation.

Related questions

Will blocking GPTBot remove my content from ChatGPT?

No. It excludes you from future training collection. Content already in a trained model is unaffected, and your presence in ChatGPT search depends on OAI-SearchBot, which is a separate crawler with a separate rule.

How do I measure traffic from AI assistants?

Imperfectly. Some arrive with a recognisable referrer; many arrive with none and are recorded as direct. A more reliable signal is checking your server logs for the AI crawler user-agents themselves, which tells you who is reading the site even when the resulting visits are invisible.

Is blocking GPTBot worth it?

That depends entirely on how you feel about training use, not on traffic. There is no traffic argument either way. Decide the licensing question and configure accordingly.

Read next

Should I block AI crawlers from my site?

An honest look at what you gain and what you give up, and why the answer is usually "block some, allow others".

Read

Why is my website not showing on ChatGPT?

The four things that decide whether ChatGPT can find, read and cite your site — and how to tell which one is stopping you.

Read

How do I check if AI crawlers can read my site?

Four tests you can run yourself in about ten minutes, in the order that finds the problem fastest.

Read

How do I fix robots.txt blocking Google?

How to read the file correctly, what the precedence rules actually are, and the three mistakes that cause almost every accidental block.

Read

All 50 guides · How every check works · API for agents