Google Tests Paying Publishers When Their Content Helps Generate AI Answers

Google is testing payments for publishers whose content helps generate answers in Gemini, AI Overviews, and AI Mode through Search Console.

Purva September 15, 2026 7 min read

Google is testing a new program that could change how publishers are compensated when their content is used in AI-generated answers.

Through an invite-only AI Contribution Pilot, Google is testing direct payments to publishers whose content significantly contributes to answers generated across Gemini, AI Overviews, and AI Mode.

The payments are being surfaced through Google Search Console.

The program was first reported in detail by Digiday and later confirmed by Google as an early-stage experiment.

But there is an important distinction: Google is not simply paying websites every time they appear as a citation in an AI answer.

The content needs to contribute meaningfully to the generation of the response itself.

How Google’s AI Contribution Pilot works?

According to information shown to publishers participating in the pilot, content can qualify when it “contributes significantly” to an AI-generated response.

This means Google may use information from a publisher’s page while constructing or grounding an AI answer.

Simply being linked after the answer has already been generated does not qualify.

The pilot currently covers:

  • Google AI Overviews
  • Google AI Mode
  • Gemini

Publishers that join the program receive an AI Contribution section inside Search Console, where they can see monthly earnings and previous earnings information.

However, Google currently provides very little detail about how those payments are calculated.

Publishers cannot clearly see which article contributed to a particular AI answer, how often their content was used, or exactly why a certain contribution resulted in a particular payment.

One publishing executive familiar with the program described the system to Digiday as a “black box.”

This is different from getting cited in AI answers

The distinction between AI visibility and AI contribution is important.

Google already provides a Generative AI Performance report in Search Console. That report allows website owners to see impressions their pages receive from AI Overviews and AI Mode.

The AI Contribution Pilot measures something different.

A website might appear as a source or link inside an AI response without qualifying for payment.

For the publisher to earn through the pilot, its content needs to materially contribute to the AI-generated response.

In other words:

Being visible in an AI answer does not necessarily mean your content helped generate that answer.

Google is beginning to treat those as two separate things.

Google has been building toward this for months

The AI Contribution Pilot did not appear overnight.

References to an “AI contribution pilot” were spotted in Search Console documentation as early as April 2026, although it wasn’t clear at the time what the feature would do.

Then, on June 18, Google publicly announced that it was testing new ways to work with websites whose content meaningfully contributes to the freshness and factual accuracy of generative AI responses through grounding.

At the time, Google did not explain that some of those partnerships would involve direct payments through Search Console.

Google has also been expanding the controls and reporting available to publishers around AI Search.

As of August 31, website owners worldwide can control whether their content is eligible to appear in and help ground Google’s generative AI Search experiences.

Google also introduced a dedicated Generative AI Performance report showing impressions from AI Overviews and AI Mode.

The AI Contribution Pilot now introduces another piece to that system: compensation.

This is not payment for training Google’s AI models

Another important distinction is that this pilot appears to focus on the use of web content when generating current AI responses.

It is not the same as Google paying publishers for using their content to train Gemini.

Google describes the process as grounding, where an AI system retrieves information from the web to make a response more current or factually accurate.

Google treats model training separately. Its Search Console documentation specifically says that the controls governing content inside generative Search experiences do not control AI training. Google-Extended is used separately for managing certain training uses.

So the AI Contribution Pilot is better understood as compensation for content helping produce AI answers, rather than compensation for training the underlying model.

How much are publishers earning?

For now, probably not much.

Industry sources cited in the original reporting described early payments as small compared with publishers’ traditional advertising revenue.

Google has also not disclosed the formula it uses to calculate the value of a publisher’s contribution.

That makes it difficult for publishers to determine whether participating in the pilot will become a meaningful new revenue stream.

The bigger development may not be the amount being paid today.

It is that Google is beginning to assign monetary value to the information websites contribute to AI-generated answers.

Why this matters for publishers and SEO

Traditional search created a fairly simple exchange.

A website published useful information. Google ranked it. A user clicked the result. The publisher then had an opportunity to monetize that visit.

AI-generated search changes that relationship.

Google can now use information from multiple websites to construct an answer before the user visits any of them.

That has raised a difficult question for publishers: what happens when your content creates value for an AI answer but does not result in a visit to your website?

The AI Contribution Pilot appears to be one experiment in answering that question.

It could also introduce a new way of thinking about AI search optimization.

Much of today’s GEO and AI visibility conversation focuses on whether a website gets cited.

But Google’s own pilot suggests another metric may eventually matter:

How much does your content actually contribute to the answer?

That could make original reporting, first-party research, unique data, expert analysis and genuinely new information even more valuable in an AI-driven search environment.

Google has not said that publishing these types of content will increase payments or improve AI visibility, so it is too early to turn this into a new optimization formula.

But the direction is worth watching.

For the first time, Google is testing a system where the value of web content inside AI-generated answers may be measured not only in impressions and clicks, but in direct compensation to the publisher.

As search becomes more AI-driven, SEO teams have more than rankings to keep track of. Search visibility, traffic changes, landing-page performance, and content performance increasingly need to be looked at together.

Vaizle brings data from Google Search Console, Google Analytics, and your website context into one workspace. SEO and marketing teams can use it to review performance, investigate changes, compare pages, and set up recurring agents for reports and checks without rebuilding the same analysis every week.

Try Vaizle

Resources You Might Find Interesting
Facebook Analytics
Improve your Facebook strategy with data-driven insights. Read blogs to grow smarter.
Explore More
Instagram Analytics
Learn how to grow on Instagram with data-driven tips. Explore blogs for latest insights.
Explore More
LinkedIn Analytics
Get better at YouTube with data-backed strategies. Read blogs to level up your channel.
Explore More
YouTube Analytics
Optimize your LinkedIn strategy with expert tips. Check out blogs for actionable insights.
Explore More
Marketing Strategies
Discover powerful marketing tactics to grow your brand. Read blogs for fresh ideas.
Explore More
Social Media Marketing
Stay ahead with the latest social media trends. Browse blogs for smart marketing tips.
Explore More