What AI assistants can read about your store, and the four things you control

Start with what Google says about its own AI features. Its guidance for AI Overviews and AI Mode says you don’t need new machine-readable files, AI text files or special schema.org markup to appear in them. So treat everything below as housekeeping. It removes reasons for a wrong or missing answer, and it does not switch citations on.

1. The AI crawler lines in robots.txt

Each AI company publishes the names of its crawlers and what each one is for.

• OpenAI lists OAI-SearchBot, which helps sites appear in ChatGPT search, and GPTBot, which collects content for training. You can allow one and block the other.
• Perplexity says PerplexityBot surfaces and links sites in its search and is not used to train foundation models.
• Anthropic lists three: ClaudeBot (training), Claude-User (fetches a page when a person asks Claude something) and Claude-SearchBot (search quality).
• Google-Extended has no crawler of its own. It is a name you can use in robots.txt as a control, while Google keeps crawling with its usual user agents.

If your robots.txt has User-agent: * and nothing that blocks your pages, all of these crawlers are already allowed. You only need a named line when you want to block one.

The vendors add and rename crawlers. Anthropic’s page, for example, now names three agents, so a robots.txt line written for one older name does not cover the others. Compare your file with the vendors’ own pages rather than with a list copied from somewhere else.

On Shopify, the platform generates robots.txt for every store. Any AI crawler the file does not name falls under its general User-agent: * rules.

Shopify’s help center says you can change the file through a robots.txt.liquid template. The same page calls that an unsupported customization that Shopify Support can’t help with, and warns that a mistake can cost you all your traffic. Not yet tested by us

2. llms.txt

llms.txt is a proposal by Jeremy Howard, published at llmstxt.org. It is a markdown file at your site’s root with a short summary and links to your important pages. It is a proposal, not a standard, and Google’s guidance above says its AI features don’t need it.

This site publishes one at boldnsmart.com/llms.txt, as an example of the format. How to place one at the root of a Shopify store is not covered here.

3. Organization markup with sameAs

Google’s documentation for Organization structured data shows a sameAs list: links to other profiles of the same organization. It tells a machine that these pages describe one company.

4. The same facts everywhere

If an assistant finds two versions of your name or description, it has to pick one. A name written three ways, for example with and without an apostrophe, gives it three candidates. Schema.org’s alternateName field lets your markup list the other spellings as names of the same thing. Pick one spelling and one short description for your store, then use them on your site, your listings and your profiles.