Will blocking AI training crawlers make ChatGPT forget my brand?

Blocking a training crawler has no measured effect on brand recall, in either direction. That is not a dodge. It is the state of the evidence.

A model like ChatGPT answers in two ways. Without web search, it relies on what it learned in training, frozen at a knowledge cutoff. With web search, it fetches pages at answer time and can cite them. Training crawlers like GPTBot only feed the first path.

So the question splits in two:

• Search answers: blocking GPTBot does not touch them. OpenAI's crawler documentation says "Each setting is independent of the others", and gives the example of allowing OAI-SearchBot while disallowing GPTBot.
• Answers from memory: in theory, training data is the only route into these. In practice, nobody has tested whether excluding one site changes what a model recalls about a brand. OpenAI does not claim it does. No independent study has measured it.

The practical upshot: you can make the training decision on its own merits (licensing, policy, preference) without assuming it costs you search visibility. For the full trade-off, see whether to block GPTBot and still appear in ChatGPT search.

OpenAI runs separate crawlers for separate jobs. As of OpenAI's bots page (verified August 2026):

A site that wants to opt out of training but stay eligible for ChatGPT search can say so in robots.txt:

User-agent: GPTBot
Disallow: /

User-agent: OAI-SearchBot
Allow: /

The most common real mistake runs the other way. A blanket block on every AI user agent stops training use and also removes ChatGPT search eligibility. If you want a definition of the category, read what a training crawler is and what blocking it does. The same pattern holds at other vendors: Anthropic separates ClaudeBot (training) from Claude-SearchBot (search).

Three things are documented. One thing is not.

Documented: GPTBot is a training crawler. OpenAI's page describes training use and asserts no search or citation benefit from allowing it.

Documented: a model's built-in knowledge has a fixed cutoff. As of August 2026, OpenAI listed a February 16, 2026 knowledge cutoff for the GPT-5.6 family and August 31, 2025 for the Instant path used in ChatGPT. A block you add today cannot reach any model already trained. Our guide to knowledge cutoffs and why AI knows your old news covers this in more depth.

Documented: brands that are prominent across the web get cited far more often. That association reflects mentions on many sites, not just your own pages.

Not documented: whether removing your own site from training data lowers recall. A July 2026 critical survey of 45 GEO studies (arXiv 2607.14035) found no technique with a stable, cross-platform causal effect on discoverability. Nothing in that literature tests training exclusion against brand recall.

What that means: anyone who says blocking GPTBot will make ChatGPT forget you, or that allowing it will make ChatGPT remember you, is guessing.

You may already be blocking training without knowing

From September 15, 2026, Cloudflare blocks Training and Agent crawlers by default on ad-displaying pages for new domains, new sites on existing accounts, and existing free-tier customers. Search crawlers stay allowed by default. Your robots.txt may say one thing while your CDN does another.

A firewall can block the crawler you meant to allow

OpenAI's help centre says eligibility requires allowing OAI-SearchBot and confirming your host or CDN accepts traffic from its published IP addresses. A 403 at the firewall removes you from search as completely as a robots.txt block. See when your CDN or firewall blocks AI crawlers.

Old crawls do not disappear

Models already trained keep what they learned. A new block only affects future collection, which is why blocking now does not make AI forget pages it already crawled.

Google works differently

Google-Extended controls training and grounding for Gemini Apps, and Google says it does not affect Search inclusion. There is no robots.txt token for AI Overviews or AI Mode. Read how Google-Extended affects Gemini's brand answers.

Make the training decision deliberately, then verify the search path stays open.

1. Decide on training separately. Choose whether to allow GPTBot, ClaudeBot and Google-Extended based on your content policy. No evidence ties that choice to recall.
2. Allow the search crawlers you want. Check that OAI-SearchBot, Claude-SearchBot and PerplexityBot are not caught by a blanket AI block. The robots.txt Checker for AI Crawlers reads each rule per bot.
3. Test the network layer. Request a page as each search crawler and confirm a 200, not a 403 or a challenge page.
4. Put key facts in the HTML. Most AI crawlers other than Googlebot do not run JavaScript, so state prices and product facts in server-rendered HTML.
5. Watch what engines say, per engine. The free checker on this site shows whether AI answers name you for up to three buyer questions.

For a step-by-step setup, follow the playbook to block AI training but stay in AI search.