Google's Quiet Licensing Pilot Covers About 100 Publishers
Google has assembled roughly 100 digital publishers into a pilot program that compensates them for material surfaced inside its AI products. Reporting from The Information indicates the arrangement covers content used in AI Overviews, AI Mode and the Gemini chatbot, and that it began less than a year ago. It is the company's most concrete attempt so far to put a price on the articles, posts and reference material its models summarize, quote and recombine into generated answers.
Participants can open Google Search Console to see how frequently their work appears in an AI answer and how much they have accrued from it. That reporting is one of the few transparent elements of the deal. What most publishers cannot see, several told The Information, is the arithmetic underneath the payments: how much weight Google assigns to any single source, why the amounts fluctuate month to month, or what standard determines that one outlet's material is worth more than another's.
The practical result is a system in which the company that consumes the content also defines its value, with little visibility for the people producing it.
Payments Range From Trivial to Seven Figures
The figures described in the report span several orders of magnitude, which suggests the program functions less like a uniform royalty pool and more like a set of individually negotiated or algorithmically assigned deals.
- For a number of small and midsize blogs and websites, payments amount to less than one-tenth of one percent of their advertising revenue.
- Some smaller sites have collected under $1,000 spread across several months of participation.
- At least one publisher reported receiving between $50,000 and $60,000 over a few months.
- Another participant earns more than $1 million per year from the program.
The gap between the smallest and largest recipients is enormous. For the outlets at the bottom of the range, the money is closer to a rounding error than a licensing fee, and it arrives against a backdrop of declining referral traffic from traditional search results.
Niche Audiences Appear to Earn More
According to the report, content covering specialized subjects with devoted followings tends to generate larger payments. Anime and gaming were cited as examples of topics that appear to command higher compensation. The pattern hints at a preference for material that AI systems cannot easily synthesize from widely available sources — or that users ask about frequently enough to justify licensing it.
Several publishers also said their payments can shift from one month to the next with no explanation attached. Without a documented methodology, recipients have no way to audit the numbers or to challenge a drop.
Some Publishers Are Refusing to Sign
Not everyone is participating. Larger publishers are reportedly declining to join in an effort to pressure Google into offering better terms. Their leverage is limited, however, because their search traffic is already eroding, and multiple studies have found that AI Overviews substantially reduce visits to the open web.
The decision facing publishers is uncomfortable. Accepting a small check locks in a low benchmark for the use of their work. Holding out preserves the possibility of a better deal but forfeits near-term revenue while traffic continues to slide.
Legal and Regulatory Pressure Is Mounting
The commercial dispute has moved well beyond private negotiations. The timeline of formal challenges has accelerated over the past two years.
- In July 2025, independent publishers filed a complaint with the European Commission over AI Overviews.
- In September 2025, Penske Media — the parent company of Rolling Stone — sued Google, citing lost traffic and advertising revenue.
- In December 2025, the European Commission opened an antitrust investigation into whether Google imposes unfair terms by using publishers' content for AI features without adequate payment or a genuine way to opt out.
A ruling in Germany adds a separate legal dimension. A German court determined that AI Overviews constitute Google's own content rather than summaries of existing material. If that interpretation spreads to other jurisdictions, publishers would gain a legal foundation to demand licensing fees whenever their work appears in a generated answer.
Why a Fair Payment System Is So Hard to Build
The economics of compensating sources inside a generative system are genuinely difficult, and that difficulty is part of why the current arrangements look the way they do.
Attributing a specific dollar value to one source's contribution is a hard problem. An AI answer may draw on fragments from dozens of documents, blending facts, phrasing and structure in ways that do not map cleanly onto any single article. Isolating the marginal value of one page within that mixture is difficult, and in many cases arguably impossible.
Even if attribution could be solved, administering it would be costly. Metering every source, calculating shares and distributing payments to thousands of sites would introduce significant overhead. And if the resulting fees were set at a level publishers consider fair, they could cut meaningfully into Google's margins.
For now, Google controls the terms on every front. It chooses which publishers receive selective licensing deals, it offers an opt-out mechanism, and it operates a payment model that lets it determine what content is worth. That combination creates a divide-and-conquer dynamic: each publisher must weigh whether a modest payment is better than nothing, while the entity setting the price also benefits from low prices.
The result is a prisoner's dilemma. Collective refusal would give publishers far more bargaining power, but any single outlet that breaks ranks gains an immediate advantage over its peers. As long as that incentive exists, the small checks keep flowing.
What to Watch Next
Three developments are likely to shape the next phase of this dispute. The first is the European Commission's antitrust investigation, which could force disclosure of how payments are calculated or impose requirements on how Google licenses publisher content for AI features.
The second is the legal question of whether AI Overviews count as Google's own output or as derived from existing material. A German court has already leaned toward the former, and wider adoption of that view would open the door to mandatory licensing fees.
The third is publisher behavior. If traffic losses continue and the pilot's payouts stay small, more large outlets may decide that withholding content is worth the short-term cost — turning a fragmented group of holdouts into something closer to a bloc. Until then, roughly 100 publishers remain inside a program whose rules, pricing and long-term value are still written almost entirely by one side of the table.
This article is based on reporting by The Decoder. Read the original article.
Originally published on the-decoder.com
