Generate an ai.txt file that states who you are and what AI agents and crawlers may and may not do with your content.
How to use the ai.txt generator
- Fill in your identity and contact.
- Choose permissions: training, summarising, citing.
- Download it and serve it at /ai.txt.
ai.txt vs robots.txt
robots.txt controls crawling. ai.txt states usage preferences for AI systems in one place. It is an emerging convention, so keep robots.txt rules as the enforceable part.
What each part of the file says
The identity block gives your name, URL, a contact address and a one-line summary, so an AI system knows who it is reading. Permissions has three switches. search-and-answer says whether answer engines may quote you; leave it allowed if you want to be cited. model-training says whether your content may be used to train models. require-attribution asks for a link or credit when you are quoted. Restrictions lists paths that should not be used at all, such as account or checkout pages. The last line points to /llms.txt for a curated map.
Making it consistent
ai.txt and robots.txt should say the same thing. If training is set to not allowed, block GPTBot, Google-Extended, CCBot and other training crawlers in robots.txt too, because that is what crawlers actually enforce. If answers are allowed, make sure OAI-SearchBot, PerplexityBot and Claude-SearchBot aren't blocked there. Use a contact address someone reads; licensing and correction requests may arrive there.