AI NewsGo-to-marketReported
The Information reports one publisher in Google's AI contribution pilot earned over a million dollars in a year and another earned 50 to 60 thousand
The Verge reports The Information found one publisher in Google's AI contribution pilot has earned more than a million dollars in a year, and another that joined months ago has earned 50 to 60 thousand, in the first per-publisher earnings figures from the pilot.

Image: The Verge
Why it mattersThese are the first per-publisher dollar figures a GEO team can quote back to a stakeholder, and the range of 50,000 to 1 million a year sets a plausible floor and ceiling for what being cited by Gemini, AI Overviews and AI Mode is worth to a single site today.
A page being cited by Gemini, AI Overviews and AI Mode has had a number next to it inside Google Search Console since 14 September, and until this week nobody outside the pilot knew what that number looked like. The Information's reporting, carried on 30 September 2026 by The Verge, puts two of those numbers on the table: one early joiner has earned over a million dollars in a year, and one publisher that joined months later has earned $50,000 to $60,000.
What the Verge piece reports
Emma Roth's piece in The Verge, dated 30 September 2026, pulls four facts from The Information's story. Google is running a pilot that pays participating publishers when their content contributes to AI-powered search features. About 100 publishers are in it. Payments cover contribution to AI Overviews, AI Mode and the Gemini chatbot. Digiday first reported the pilot, which Google confirmed on 14 September; The Information adds the per-publisher earnings range on 30 September. The two named numbers are the only two the piece quotes.
The size of the gap, and what it means for a per-site estimate
A $50,000 to $60,000 figure and a $1 million figure, on the same programme, 100 sites in the pool, is a range of roughly 20 to one. On its face that is not a flat per-citation rate: the gap is wide enough that the amount a page earns depends on something other than being cited, likely the number of answers it feeds and how heavily it is used in each one. Neither The Verge nor The Information publishes a per-page or per-citation rate, so a site using the pilot's own Search Console widget sees a monthly total it cannot break apart.
Where this sits against earlier coverage of the same pilot
The pilot's own structure has been public for two weeks. The AI contribution panel appears inside Google Search Console on invite, shows a monthly earnings figure, and credits pages that "contribute significantly" to an AI-generated answer. A page cited by an AI response after it is written does not qualify, Google's help text says. Early executive quotes to Digiday called the payouts a token amount. The Information's two figures are the first that put a scale against that word. One million a year, on the top end, is a number a mid-sized publisher's finance team would plan around; $50,000, on the lower end, is a number a GEO lead can put on a slide.
Regulators are watching the same programme from a different direction. The UK's June 2026 ruling said publishers must be allowed to opt out of appearing in AI search features, and the European Union has opened an investigation into Google's search traffic impact. For a team whose own site is already feeding Gemini, AI Overviews and AI Mode without an invite, the two figures are the first dollar numbers a stakeholder can be shown. Worth watching: whether Google publishes the formula behind the monthly figure, and whether the invite list widens past the small and mid-sized publishers it reaches today.
Source
- Google reportedly tests paying publishers for AI search results, The Verge, Emma Roth, 30 September 2026
- Original reporting: Why Google Is Paying About 100 Digital Publishers for AI Overviews, The Information (paywalled)
This item was written by an AI system from the linked source. Reveneau is responsible for what it publishes.
Get AI News in your inbox
New developer tools, model and agent releases, and how teams are actually using them to release software. Short, and only when there is something worth reading.


