Google lets Gemini Notebook read your bought books and turn them into AI research tools

Google has expanded Gemini Notebook so users can pull in books they’ve bought from Google Play Books and interrogate them with AI, turning personal libraries into searchable, interactive research spaces. The new feature, called Expert Intelligence, matters because it gives readers, students and professionals a way to ask questions about purchased titles and generate summaries, plans, infographics and other outputs from their contents.

The rollout marks another step in Google’s effort to make Notebook-based AI a central part of its broader ecosystem, while also creating a framework that keeps book access tied to ownership and licensing. More than 100,000 titles are supported at launch, and Google says it plans to widen the feature beyond books to include scholarly work, magazines and newspapers.

Gemini Notebook, Google’s AI note-taking and research app, is becoming more than a place to store links and notes. With Expert Intelligence, it now acts like a conversational layer over books people already own, letting them ask follow-up questions, build study guides and produce custom materials based on the text. Google is also positioning the feature as a possible new revenue channel for authors and publishers.

What Google announced

Google is adding a new book-based capability to Gemini Notebook that lets users connect purchased Google Play Books titles to the app and use them as source material for AI tasks. In practical terms, that means a reader can import an eligible book into Notebook and then ask Gemini questions about it, synthesize ideas from it, or generate derivative materials such as outlines, podcasts, visuals and planning documents.

The feature is branded Expert Intelligence inside Notebook. Google’s pitch is that it helps users build a private, personalized reference library of authoritative sources instead of relying only on web search or generic AI responses.

How Expert Intelligence works

Expert Intelligence works by linking titles from Google Play Books directly into Gemini Notebook. Once a book is added, it can be used like any other source within the app, including for direct Q&A and for creating new content formats from the material.

Google says users can also open the full book text inside Notebook, which makes the experience closer to an AI-enhanced reading environment than a simple citation tool. That could be especially useful for students, writers and managers who need to extract specific lessons from dense texts without constantly switching between apps.

Google Labs editorial director Steven Johnson said the idea is to help people assemble a personal “dream team” of trusted experts whose work can be consulted wherever they are.

Why this feature matters for readers and publishers

The biggest significance of the update is that it connects AI assistance to books people have legitimately purchased. That changes Notebook from a general-purpose productivity tool into a more opinionated knowledge platform built around licensed content.

For readers, the appeal is obvious: instead of manually skimming chapters or flipping between notes, they can ask direct questions, compare themes across books and generate tailored outputs from the source material. For authors and publishers, the feature could create new demand for backlist titles and premium educational content if AI-assisted reading becomes a common habit.

Google has framed the launch in commercial terms as well, saying it wants to create a setting that can support new revenue opportunities for rights holders. That matters at a time when publishers are pressing AI companies to respect copyright, pay for licensed material and avoid training or retrieval systems that bypass the normal economics of publishing.

How Google is trying to avoid access problems

Google says it has built safeguards so people cannot view or use the contents of books they did not purchase. If a Notebook is shared with another user, that person may see the book listed as a source, but they will not be able to read the full text or access information derived from it unless they own the title themselves.

When a non-owner clicks through, Google sends them to Google Play Books so they can buy the book. That design is important because it turns the feature into a commerce funnel as much as a productivity tool.

In other words, Expert Intelligence is not meant to make books freely unlockable through AI. It is intended to preserve the link between ownership and access while still enabling AI-driven interaction for the buyer.

What kinds of books are supported at launch?

At launch, Google says the feature works with more than 100,000 books from a broad list of publishers, including Penguin Random House, Johns Hopkins University Press, Macmillan Publishers and O’Reilly Media. The range suggests the company is targeting both mainstream trade readers and more specialized professional and educational audiences.

The initial catalogue gives the feature immediate depth, but it also highlights that Google is relying on publisher relationships to make the product useful. Without broad licensing, an AI book tool would quickly hit the limits of its own library.

How Google is using authors to shape the product

Google is not limiting the launch to publisher catalogs. The company also worked with 15 authors to create Featured Notebooks, which are customized collections that include extra source material, introductory letters and related context chosen by the writers themselves.

Among the authors involved are Michael Pollan, Kim Scott, Steven Pinker and Jennifer Wallace. Their participation gives Google a way to showcase not just books, but guided knowledge experiences built around the people who wrote them.

That approach could be especially valuable for nonfiction, where readers often want interpretation, context and practical takeaways rather than just raw text. A guided Notebook can turn a book into a more active learning environment, with the author effectively serving as curator and host.

Johnson said Google wants to create a place where users can bring in the knowledge they think is most relevant while also opening the door to new revenue possibilities for authors and publishers.

Examples Google demonstrated

During a briefing, Google showed how the feature could work in everyday use. One demonstration used Michael Pollan’s Food Rules to help generate a recipe book, while another used Kim Scott’s Radical Candor to answer a management question about handling a difficult conversation with an employee.

Those examples point to the real target audience: people who are trying to turn reading into action. Google appears to be aiming at users who want not just summaries, but practical outputs they can apply immediately.

• Study prompts based on a textbook or academic work
• Meeting plans informed by a business book
• Infographics built from a nonfiction title
• Audio-style AI podcasts summarizing a book
• Custom research briefs assembled from several sources

Why this is a bigger move than a simple app update

The launch is more than a feature tweak because it shows how Google intends to blur the boundary between reading, searching and generative AI. Notebook is increasingly becoming a place where information is not just stored but actively reworked into new forms.

That has major implications for digital publishing and knowledge work. If the model catches on, readers may begin to expect every serious book to come with a machine-readable companion experience. Publishers could then find themselves negotiating not only for sales, but for placement inside AI workflows.

It also gives Google a stronger answer to competing AI tools. Many chatbot products can summarize text pasted into a prompt, but few are designed to manage licensed books, preserve ownership rules and link directly to a commercial reading platform. That combination is strategically important.

How it fits into Google’s broader AI strategy

Google says the feature will eventually extend beyond Gemini Notebook into AI Mode in Search and the Gemini app. Both already draw from Notebook in some capacity, which means the company is building a connected ecosystem where source material can move across products.

That ecosystem strategy is familiar for Google. The company tends to strengthen one product by making it feed or reinforce another. In this case, books purchased in Play Books can enhance Notebook, which can then power Search and Gemini.

Google also plans to add support for scholarly articles, magazine articles and newspapers. That would expand the system from a book companion into a broader research layer capable of serving academics, journalists and knowledge workers.

What the feature means for copyright and ownership

The rights question is at the center of this launch. AI companies have faced growing pressure over how they handle copyrighted books, journalism and other licensed material. Google’s answer here is to make the feature opt-in through ownership and to limit what non-buyers can do with shared notebooks.

That design is unlikely to end broader copyright debates, but it does show a different model from scraping or indiscriminate text ingestion. Instead of treating books as a universal training set, Google is tying access to the user’s purchase record and to the publisher ecosystem.

For publishers, that could be an attractive compromise if it increases sales and keeps control over access. For users, it means a more controlled and potentially more trustworthy AI reading tool. The unanswered question is whether consumers will see enough value to make this kind of linked ownership model mainstream.

How Google’s new book AI compares with traditional note tools

Expert Intelligence is different from ordinary note-taking because it does not just organize your thoughts. It also lets the content itself become a conversational source that can generate new outputs on demand.

Traditional note tools help users file away excerpts or manually tag key passages. Gemini Notebook goes further by letting the system interpret the source and build on it. That means the app is less of a filing cabinet and more of a working assistant.

For professionals, this could be especially useful in fields where books remain core reference material. Managers, educators, researchers and consultants may be able to turn a few well-chosen titles into a personalized advisory system.

What happens next

Google has signaled that the book feature is only the beginning. The company’s stated roadmap includes bringing Expert Intelligence into more parts of its AI stack and adding a wider range of source types.

If those plans materialize, users could eventually build notebooks around journals, periodicals and news coverage in addition to books. That would make Gemini Notebook a more serious competitor to research databases and premium reading tools, not just a companion for casual readers.

The success of the feature will likely depend on three things: how well the AI handles source material, whether publishers embrace the licensing model and whether users see enough practical value to make the app part of their regular reading habits.

For now, Google is betting that people want an AI layer on top of the books they already own. If that bet pays off, Gemini Notebook could become a central hub for individualized learning, professional development and content discovery across the broader Google ecosystem.

Key facts at a glance

• Google has added book-based AI features to Gemini Notebook through Expert Intelligence.
• The feature works with purchased titles from Google Play Books.
• More than 100,000 books are supported at launch.
• Google says it has protections to stop non-owners from accessing full text in shared notebooks.
• The company plans to expand support to scholarly articles, magazines and newspapers.

Why this launch may matter beyond Google users

Google’s update is part of a larger shift in AI products from generic chat toward curated, licensed and source-aware systems. If the model becomes popular, it could influence how other companies build research assistants, educational tools and reading platforms.

It also reflects a growing recognition that AI tools are more useful when they are grounded in material people already trust and pay for. Books remain one of the most durable knowledge formats, and Google is trying to make them more interactive without stripping away the business model that supports them.

That balance between convenience, rights protection and commercial opportunity will likely define the next phase of AI publishing products. Google’s latest move suggests it wants to be the company that sets that standard.

In the short term, the update makes Gemini Notebook more capable. In the longer term, it could help determine how much of the world’s paid knowledge becomes part of everyday AI use.

And that is why this launch matters: it is not just about taking notes on books, but about redefining what ownership, access and assistance look like in an AI-first reading environment.