1. AI cites, but does not pay: the broken economy of the citation

For twenty years, the web ran on an implicit contract. A search engine indexed your content, and in exchange it sent you traffic. You turned that traffic into revenue: ads, subscriptions, sales. The citation was worth a click, and the click was worth money.

AI answer engines broke that contract. When a user asks ChatGPT, Perplexity, or Gemini a question, the AI reads your articles, extracts the substance, and answers directly. The answer is complete. The user no longer needs to click. Your content was used, your source may even have been cited, but no traffic and no revenue came back to you.

This is not a theoretical problem. It is already measurable, as long as you look in the right place. AI bot traffic is invisible in Google Analytics, which relies on JavaScript that most AI crawlers do not execute. To see it, you have to measure at the server level. We break this down in our article on the AI traffic Google Analytics cannot see.

2. The new wave: how AI can pay the sources it cites

The good news is that the topic is now on the table, and several concrete models are emerging to move value toward sources. None is the standard yet, but together they sketch a citation economy being built in real time.

Direct licensing deals

The major AI providers sign licensing deals with large publishers to use their content. It works, but it is reserved for a handful of players with the leverage to negotiate. The French press is organizing a collective response, which we analyze in our article on the press alliance facing AI licensing.

Revenue-share programs

Some answer engines, Perplexity in the lead, have launched programs that share part of their revenue with the publishers whose content is cited. The idea looks like an affiliate model applied to the citation.

Pay-per-crawl

Infrastructure players such as Cloudflare have introduced pay-per-crawl: charging for access to content at the moment a bot fetches it. It is the logic of a toll rather than a wall. The x402 protocol pushes this idea even further by turning every HTTP request into a payment opportunity, as we explain in our article on the x402 protocol.

Citation-based revenue attribution

A more granular approach measures which sources actually contributed to an AI answer, and distributes compensation based on that contribution. Several young companies are building this attribution model.

New-generation ad networks

And then there is an approach coming from advertising, which funds the ecosystem through an entirely new channel rather than through negotiation. That is exactly the bet uwait is making.

3. uwait: "Get paid while AI thinks"

uwait starts from a simple observation. When you ask an AI a question, there is dead time: those few seconds while the model thinks and generates its answer. That waiting time has no value today. uwait proposes to turn it into value, and to redistribute it.

How it works

1. The user asks a question to their AI assistant (ChatGPT, Claude, Perplexity, Gemini, Mistral, Copilot, DeepSeek, Grok) through a browser extension.
2. A real-time auction fires while the AI generates its answer.
3. A native ad appears discreetly, below the answer, without interrupting the experience.
4. The user is paid immediately for their attention.
5. The revenue is split across all stakeholders.

The detail that matters here is those 30% paid to cited sources. For the first time, simply being cited in an AI answer triggers a flow of revenue toward the publisher, without any negotiation required. The citation becomes monetizable again, through an unexpected path: advertising on the user side rather than traffic on the publisher side.

4. Why paying 30% to sources changes the game

Most source-compensation models rest on leverage. To sign a license, you need a catalog big enough to interest an AI provider. To negotiate, you need a brand strong enough to matter. The result: very large publishers capture value, and the rest of the web is left behind.

The appeal of a model like uwait's is that it depends on no negotiation. If your content is cited in an answer, you are eligible for your share. The small publisher, the niche blog, the specialized outlet enter the same loop as the giants. It is an ecosystem logic rather than a power game.

None of these models will win alone. The citation economy will probably be a stack: a license for large catalogs, a toll for crawlers, a revenue share for visibility, and an advertising channel to fund mainstream usage. What matters is that every publisher can plug into these flows as they come online.

5. Where Senthor fits in this economy

All these models share one prerequisite that often goes unsaid: you first have to know that AI is using your content. You cannot monetize, negotiate, or regulate traffic you cannot see. And that traffic is precisely invisible in standard tools.

That is Senthor's role. Installed at the server level or on the edge, Senthor intercepts every request before it reaches your application, and identifies AI bots even when they pretend to be regular browsers. You then get three things:

If you run a site and want to know where you stand, start with the simplest step: measure. You can estimate what your AI traffic represents with our revenue estimator, or learn how to monetize bot traffic instead of blocking it.

FAQ: paying the sources AI cites