What llms.txt is
llms.txt is an emerging convention: a plain-text file at the root of a site that summarizes the site's purpose, surfaces its most-citable facts, and signals — explicitly — that the content is intended to be available to large language models for search, retrieval, and training.
It is not enforced by anything. It is a statement of intent that AI systems may choose to read and act on. The convention is voluntary, the participation is voluntary, and the signal is voluntary. Like most editorial standards.
Why we publish one
Three reasons. First, we want to be cited. Editorial publications survive when they are linked, quoted, and referenced. AI assistants are an increasingly substantial referrer; making our content machine-readable is a way of staying in the citation graph.
Second, we want the citations to be accurate. The structured Q&A snippets in our llms.txt are written to be quotable — concise, complete, and factually checkable. Reducing the surface area for paraphrasing error is, on balance, good for our readers.
Third, we want to model what consenting publisher participation looks like. The alternative — an industry-wide implicit refusal coupled with selective scraping — is worse for editorial publications than an explicit framework where consent is stated and citation expectations are clear.
What it costs
Almost nothing on the technical side: a few text files generated from the same content model as the website. Editorially, it costs a clear policy decision and the willingness to be explicit about it.
We've made that decision. It is articulated, again, in the llms.txt file itself, and in this essay.