Google has transformed Gemini Notebook into a paid-knowledge distribution channel by integrating over 100,000 books from Google Play Books, allowing users to query, summarize, and generate AI-powered content from titles they own. The feature, called Expert Intelligence, launched on August 27, 2026, and represents a fundamentally different approach to how AI companies license published content compared to the broad training deals that have dominated the industry.

What Makes Google's Book Integration Different From Other AI Licensing Models?

The structural innovation here matters more than the feature itself. Unlike OpenAI and Anthropic, which have pursued broad training agreements and retrieval from open web sources, Google is binding a specific, paid, per-user content library to a specific product surface. When a user shares a notebook built on a purchased book, collaborators who don't own the title see it listed as a source but cannot view the underlying text or generate content from it. Clicking the blocked source routes them to Google Play Books to buy the title. This mechanic converts each shared notebook into a soft storefront, embedding the revenue opportunity directly into the product rather than negotiating it as a separate licensing pool.

Publishers including Penguin Random House, Macmillan Publishers, O'Reilly Media, and Johns Hopkins University Press are participating at launch, spanning trade nonfiction, tech reference, and academic press. This mix signals Google is pitching the feature at both consumer readers and professional users who buy reference material.

How Can Users Actually Use Books as AI Sources in Gemini Notebook?

The mechanics are straightforward and practical. A user who owns a title on Google Play Books can pull it into Gemini Notebook as a source, then ask questions, generate infographics, plan projects, or spin up an AI-narrated podcast about the material. Google Labs editorial director Steven Johnson demonstrated using Michael Pollan's "Food Rules" to generate a recipe book, and Kim Scott's "Radical Candor" to draft an answer to a management question about a difficult employee conversation.

Supported titles are flagged with a Gemini Notebook badge under the Tools section in Google Play Books, making it easy to identify which books work with the feature. Google worked with 15 authors, including Michael Pollan, Kim Scott, Steven Pinker, and Jennifer Wallace, to build Featured Notebooks with custom source material and intro letters from the authors themselves.

Steps to Build Your Own AI-Powered Knowledge Base

• Identify Supported Books: Look for the Gemini Notebook badge in Google Play Books under the Tools section to find titles that work with Expert Intelligence.
• Import Your Sources: Pull purchased books into Gemini Notebook as queryable sources, creating a personalized collection of expertise you value.
• Generate Custom Content: Ask questions, create infographics, plan projects, or generate AI-narrated podcasts based on your imported books.
• Share Strategically: Share notebooks with collaborators, knowing that non-owners will see sources but cannot access text without purchasing the title themselves.

"I'm really interested in the idea of building a collection of people whose work you really value, and kind of creating your own dream team of advisors that are just available to you wherever you go," said Steven Johnson, Google Labs editorial director.

Steven Johnson, Editorial Director at Google Labs

Why Is This Revenue Model Significant for the Publishing Industry?

Publishers have spent the past three years suing and negotiating with AI model providers over training-data use. Expert Intelligence represents a different economic story. Rather than paying a lump sum for corpus rights, publishers get a per-title purchase requirement tied to each end user who wants the AI-assisted experience. This creates an incentive structure where publishers benefit directly from the feature's adoption and sharing.

Google framed the pitch around personalized advisor stacks rather than generic search. The company stated it wants to "create a space where users can bring whatever knowledge, whatever expertise they think is going to be most relevant and do it in a way that creates new revenue possibilities for authors and publishers as well".

What's Next for Expert Intelligence?

The launch catalog of 100,000 books is a fraction of Google Play Books' full inventory, and coverage will depend on how quickly other publishers opt in. Google plans to extend Expert Intelligence beyond books to scholarly articles, magazine articles, and newspapers, and the feature will eventually appear in AI Mode in Search and the Gemini app, both of which already pull from Gemini Notebook.

One significant caveat remains: it is unclear how the system handles quotations, paraphrase, and derivative outputs. A recipe book generated from "Food Rules" is transformative in some readings and infringing in others, and Google has not published the guardrails around output length or verbatim reproduction. Authors outside the Featured Notebooks program have no direct customization channel yet.

For the broader AI market, the structural move is what matters most. If Expert Intelligence expands to journals and newspapers as planned, Gemini Notebook becomes a distribution channel for licensed knowledge in a way that no other consumer AI product currently offers, and one where the publisher's incentive to participate grows with each notebook a user shares.