The deals you read about are real, but they are not the market most publishers will ever see. A handful of brand-name newsrooms have signed agreements worth serious money with AI companies. For everyone else the honest answer is that licensing pays little or nothing, because the deals are bilateral, negotiated one at a time, and reserved for publishers with the recognition and legal leverage to demand them. Independent analysis through 2025 and 2026 expects most publishers to see no meaningful licensing revenue, and warns that the same platforms eroding publisher traffic now control the licensing terms. So the practical question is not how to land a great licensing deal. It is what to do when you will never be offered one.
What an AI licensing deal actually is
An AI content licensing deal grants an AI company the right to use a publisher's content, either to train a model or to display and cite it in answers, in exchange for payment. Payment structures vary. Some deals are flat annual fees, some are per-use or per-query rates, and some are revenue shares tied to a product such as a subscription assistant. The largest deals have reportedly been worth tens of millions of dollars over multiple years to individual publishers.
Why only the largest publishers get paid
Licensing deals are negotiated one at a time. An AI company has limited incentive to negotiate with a small site when it can sign a single agreement covering a large, recognised content library. The result is a concentrated market. Analysts describe three tiers: a handful of large bilateral deals between major AI companies and major publishers, an emerging middle layer of licensing marketplaces and intermediaries, and an uncompensated majority of publishers and creators sitting outside any deal. If you are not in the first tier, the question is not how good your licensing deal is. It is whether you will ever be offered one.
The double bind
The deeper problem is who sets the terms. The same platforms whose AI products are reducing referral traffic to publishers are also the ones building and controlling the licensing infrastructure that is supposed to replace that revenue. A 2026 report described this as a double bind: publishers are losing their traffic-based income and are being offered replacement income on terms dictated by the companies causing the loss. In some cases, access for model training has been positioned as a condition of remaining visible in search at all, which makes a free choice look less free.
What licensing leaves on the table
Even a publisher with a licensing deal does not capture all of its AI value. Licensing typically covers training access or bulk display rights. It does not monetise the moment-by-moment behaviour of Live Search Agents, the bots that ChatGPT, Perplexity, Claude, and Gemini dispatch to read a specific page in order to answer a live user question. Those retrievals are the fastest-growing category of AI traffic and the closest to the point of purchase, because the user is actively researching a decision when the agent fetches the page. A flat licensing fee does not vary with that intent, and it does not pay per high-value retrieval.
The alternative for the uncompensated majority
For the publishers who will never sign a bilateral deal, the practical route to AI revenue is to monetise the read directly. Because Live Search Agents parse text and skip JavaScript, the monetisation has to live in the content layer rather than in the display-ad stack. blankspace operates here. When a Live Search Agent retrieves an eligible page, blankspace runs a real-time auction at the CDN edge and inserts a relevant, accurate brand fact into the content as editorial context, then attributes the revenue the same day. This does not require a negotiation with an AI company, it does not depend on being large enough to be noticed, and it scales to any publisher regardless of size. It is the model designed for the majority that licensing excludes.
How to decide whether to pursue licensing
If you are a large publisher with a recognised brand and a substantial content library, a licensing conversation is worth having, and it can sit alongside other monetisation. If you are everyone else, treat licensing as unlikely and build a strategy around the AI traffic you can monetise without a deal. In both cases, the first step is the same: get visibility into your AI traffic so you know how much of it is training crawlers, which a licence might cover, and how much is Live Search retrieval, which it will not.
Frequently asked questions
How much do AI licensing deals pay?
The largest reported deals are worth tens of millions of dollars to individual major publishers over several years. The typical publisher is offered nothing, and analysts expect most publishers to see no meaningful licensing revenue.
Can a small or independent publisher get an AI licensing deal?
Rarely. Deals are negotiated individually and favour large, recognised content libraries. Licensing marketplaces are emerging to aggregate smaller publishers, but they take a cut and do not guarantee meaningful payment.
Do licensing deals stop AI companies using my content for free?
Not necessarily. The rate of AI bots bypassing voluntary access restrictions has been rising, and a licence with one company does not bind another. Licensing is a commercial agreement, not a technical control.
What is the difference between licensing and pay-per-crawl?
Licensing is a negotiated agreement granting broad usage rights, usually to a large publisher. Pay-per-crawl is an automated charge applied to crawlers at the point of access, available to any publisher through an intermediary. Neither monetises anonymous Live Search retrievals well.
If licensing will not pay me, what will?
Content-layer advertising monetises the retrievals themselves. It does not require a deal with an AI company and works for publishers of any size, which is why it is the realistic option for the majority that licensing excludes.