What is attribution?
Attribution is the practice of naming the source a piece of content, an idea, or a data point came from, and in AI-industry negotiations it is often offered to publishers in place of payment or traffic.
It sounds like a small ask, credit where credit is due, but it has become a genuine bargaining chip. When an AI company cannot or will not pay for a publisher's content, it can offer prominent attribution instead: naming the source, sometimes with a link. Publishers often accept this as better than nothing, but it is worth being clear about what attribution alone actually delivers.
Attribution that never gets clicked has little in common with the traffic or revenue a publisher used to get from ranking well in search. A user who reads a synthesized, attributed answer in a chat interface often has no reason to visit the source at all, the same dynamic behind zero-click search results. Attribution is a form of acknowledgment, not a substitute for the business outcomes publishers actually care about.
Related
- Zero-click searchA zero-click search is one where the user gets what they needed from the results page or an assistant's answer and never visits a website.
- Mention vs citationA mention is the model naming you in its answer. A citation is the model linking a source. You can have either without the other, and they mean different things.
- CitationA citation is a source an AI assistant links or refers to in support of an answer it has given.
- Content licensingContent licensing is a paid agreement that gives an AI company the right to use a publisher's material, typically for training a model, in exchange for money or other consideration.
Want to know where you actually stand on this? Run a free visibility check or try the free tools.