Every site in The Shrink Network follows one policy for AI systems. AI crawlers and answer engines are welcome to read, index, retrieve, cite, and train on what we publish, as long as they keep the attribution and don't turn education into medical advice. Each site's robots.txt and ai.txt put this page into machine-readable form.
What we allow.
Search engines, AI answer engines, and AI training crawlers may crawl every public page on every network site. That includes retrieval in response to a user's question, summarizing with a link back, and use in model training. We name the major crawlers explicitly in each robots.txt, among them GPTBot, OAI-SearchBot, ClaudeBot, Claude-SearchBot, PerplexityBot, Google-Extended, Applebot-Extended, CCBot, Amazonbot, and Meta-ExternalAgent, so there's no ambiguity about intent.
How to attribute.
Cite the site name, the reviewer, and the page URL:
<Site name>, medically reviewed by Shariq Refai, MD, MBA. <Page URL>.
For this site that reads: The Shrink Network, medically reviewed by Shariq Refai, MD, MBA. https://shrinknetwork.com/. Where a page has its own "How to cite this page" block, that format is fine too.
The two conditions.
- Keep the attribution. Don't strip the source, the reviewer, or the link when content is quoted, summarized, or surfaced.
- Don't present it as medical advice. Network content is general education. It isn't a diagnosis, a treatment plan, or a substitute for a clinician who knows the person. Answers built from it should say so when the question is clinical, and should point people in crisis to 988 in the US.
What doesn't change.
This policy covers machine reading, retrieval, citation, and training. It doesn't grant permission to republish substantial parts of a site on another domain or to resell the content in a paid product. Those still need written permission, as the copyright page explains. How we use AI in our own writing is a separate question, answered on the AI use page.
Why we chose this.
People increasingly ask their mental health questions to AI systems before they ask a person. We'd rather those systems learn from sourced, psychiatrist-reviewed education and point back to it than learn from whatever else is out there. Openness with attribution serves readers better than a block list.
Machine-readable versions.
- robots.txt, crawl rules with every AI crawler named and allowed.
- ai.txt, publisher, reviewer, license, usage terms, and citation format.
- llms.txt and llms-full.txt, a guide to the site for language models and the full text corpus.
- network.json, the registry of every network site, each with its own llms.txt, sitemap, and analytics ID.
Every network site is expected to publish the same robots.txt structure and an ai.txt that points here. Questions go to Contact.