Google pays about 100 publishers for AI answers, and the numbers are tiny
Google is paying roughly 100 digital publishers when their content feeds its AI answers. For several small and mid-sized sites, the money comes to less than a tenth of one percent of their advertising revenue, The Information reported on Tuesday.

The figures come via The Information, which spoke to people familiar with the tests. They land badly. One publisher that joined early is making more than $1m a year. A more recent entrant has earned $50,000 to $60,000. Several smaller sites have received less than $1,000 over several months.
TNW reported the numbers on Wednesday and said it had not independently verified them. The pilot covers AI Overviews, AI Mode and Gemini. Google pays according to how much each source contributes to an answer. Sites covering niche subjects with strong followings, such as anime and gaming, tend to earn more, publishers told The Information. Several larger publishers have declined to join, hoping that staying out will push Google towards a better offer.
That is the news peg, and it is only one day old. But the story underneath it is older and considerably larger: search traffic to publishers has been falling for years, and AI answers are accelerating the fall.
What the pilot actually is
Digiday first reported the scheme on 14 September, calling it the AI contribution pilot. Jessica Davies and Sara Guaglione wrote that Google pays when a publisher's content "significantly" contributes to an AI response. Google confirmed to Digiday that the scheme is an early-stage pilot. It is testing how best to reward high-quality content and framed the payments as coming on top of the traffic and tools it already provides.
Publishers in the pilot told The Information the payments are hard to follow. Earnings appear each month in Google Search Console, but it is not clear what counts as a meaningful contribution to an answer. In other words, the people being paid cannot audit the formula that pays them.
The Decoder, reporting the same story on Wednesday, put the spread bluntly: some small and midsize blogs and websites receive less than 0.1 percent of ad revenue. One publisher got $50,000 to $60,000 over a few months. Another earns more than $1m a year. Small sites got less than $1,000 over several months. The Decoder also noted that payments can change from month to month without explanation.
So the headline number, about 100 publishers, is real. So is the seven-figure outlier. The median experience is closer to nothing.
Why publishers are not simply walking away
Because walking away has not worked. Larger publishers that refused the pilot are betting that withholding content will force a better offer, according to The Information. The Decoder frames the same dynamic as a prisoner's dilemma: a few benefit, most get little or nothing, and publishers that leave put little pressure on Google because other sources fill the gap. Individual deals keep publishers negotiating separately rather than collectively.
Meanwhile the traffic side keeps getting worse. Multiple studies show AI Overviews sharply reduce visits to the open web, The Decoder noted. Yahoo CEO Jim Lanzone said last week that micropayments will never be enough for publishers, according to TNW.
That combination, falling referrals plus token payments, is what makes the pilot politically explosive rather than merely disappointing.
The regulatory flank
Regulators are already moving. In June, the UK's competition regulator ruled that Google must let publishers opt out of its AI search features, TNW reported. The European Commission is investigating the effect of those features on web traffic and has been asking publishers whether Google's opt-out actually works. The Decoder dates the Commission's antitrust investigation to December 2025. It is examining whether Google imposes unfair terms by using publishers' content for AI features without adequate payment or a genuine way to opt out.
Litigation is running in parallel. Penske Media, which owns Rolling Stone and Variety, is suing Google over AI Overviews, per TNW. The Decoder notes that independent publishers filed a complaint with the European Commission over AI Overviews in July 2025. Penske sued that September over lost traffic and ad revenue. The Decoder also flags a German court ruling that AI Overviews are Google's own content rather than summaries of existing material. If that interpretation spreads, publishers would have a legal basis to demand licensing fees whenever their work appears in an AI answer.
Google's exposure there is not trivial. Tracking each source's exact contribution to an answer would be hard, possibly impossible. Administering the payments adds overhead. The fees themselves could cut into margins, The Decoder argues.
Publishers are also lobbying Washington. This week, executives from more than 300 news publishers lobbied Congress for a bill that would ban AI stealth bots, TNW reported. In July, Cloudflare gave AI crawlers a September deadline to pay publishers or face a block.
"The current data exchange between suppliers and supply chains involves weeks of exchanging PDFs," said Janis Vavere, co-founder of the Latvian engineering software company Trace.Space, in a separate Tech.eu interview published on Thursday.
That quote has nothing to do with publishing economics. It is included here to make a narrower point about the week's news cycle: on the same day Google's publisher payments were being dissected, Tech.eu was reporting that Trace.Space had launched Trinity. It is a hardware engineering platform that connects requirements, testing, design parameters and product variants, and gives AI agents access to that data. The company says Trinity covers around 40 percent of the product development lifecycle today, up from roughly 10 percent when it started, and aims for as much as 80 percent.
The money is moving somewhere else
Which is worth stating plainly. On Wednesday, Flow Engineering, a three-year-old San Francisco startup that uses AI agents to align CAD drawings with product requirements and simulation results, announced a $50m Series B at a $750m valuation, TechCrunch reported. The round was co-led by Antonio Gracias of Valar Equity Partners and Gavin Baker of Atreides Management. Sequoia Capital and former Sequoia partner Roelof Botha participated. Botha joined the board.
Flow names Anduril, Rivian, Joby Aviation, General Motors PPU, RV Tech and Stoke Space as customers.
The same day, OpenAI and Synopsys said they had signed a multi-year partnership to build GPT-Synopsys. It is a model for chip design that will run on OpenAI's infrastructure, with revenue shared between the two companies, according to The Decoder. Synopsys CEO Sassine Ghazi said AI could significantly speed up design work. OpenAI co-founder Greg Brockman framed the deal as a path to better chips and better AI.
None of this is publisher money. It is the other half of the same shift: capital and product effort flowing into AI systems that consume structured engineering data, while the unstructured web content those systems also ingest gets paid for at a rate publishers describe as rounding error.
What to watch
Three things. First, whether the pilot's terms change: Google has not published a formula, and publishers say they cannot reconstruct one from Search Console. Second, whether the European Commission's investigation produces anything enforceable on opt-out, which is the mechanism UK regulators already required in June. Third, whether the German ruling on AI Overviews as Google's own content survives appeal, because that is the difference between a micropayment regime and a licensing regime.
For now the arithmetic is simple. About 100 publishers are in. Several small ones have made less than $1,000 over several months. The ones staying out are betting that is a negotiating position. The ones staying in are betting it is the ceiling.
Sources
5- 01Google is paying 100 publishers for AI answers, The Information reportsEN
- 02Google is paying almost no publishers almost nothing for content used in AI answersEN
- 03Trace.Space launches Trinity to tackle hardware engineering's next bottleneckEN
- 04Valor, Atreides, and Sequoia back AI startup Flow Engineering at $750M valuationEN
- 05OpenAI and Synopsys team up to build an AI model that designs chips like a seasoned engineerEN
All figures and quotations in this text come from the sources listed below.
Content prepared by the editorial team with AI assistance.
Comments
0- No comments yet — be the first.