Google's "Expert Intelligence": Licensed Publisher Books Become the Notebook Layer's Trust Moat
On August 27, 2026, Google launched Expert Intelligence, "a cross Google initiative that helps you engage with your trusted sources through Google AI products," starting in Gemini Notebook (the product formerly NotebookLM, NotebookLM Is Now "Gemini Notebook": Google's Source-Grounded Notebook Layer, Rebranded and Enterprise-Priced). It launches "with more than 100,000 books from major publishers, including Bloomsbury, De Gruyter Brill, Johns Hopkins University Press, Macmillan Publishers, O'Reilly Media, and Penguin Random House," plus curated Featured Notebooks from "more than 15 bestselling authors — including Steven Pinker [and] Michael Pollan."
The mechanism is ownership-gated grounding: users add select ebooks they purchased on Google Play Books into a notebook and get responses "grounded directly in that book," with "direct citations to relevant content." Google built the entitlement into sharing itself: "for people to interact with a book in Gemini Notebook, they must own the book through Google Play Books. If a notebook is shared, other collaborators will be prompted to purchase their own copy to engage with the content further."
The market-map signal is the roadmap: "We're starting with books in Gemini Notebook but in the future we'll provide more opportunities to engage with Expert Intelligence-enabled content, including in the Gemini app and AI Mode in Search. Soon, people will also be able to add more sources like third-party subscriptions, business research reports, and textbooks."
What this means for the map
- The host layer is building a licensed-source supply chain. Google's answer to the trust/provenance anxiety (cf. AI Is Breaking the Proxies for Expertise: Goodhart's Law Comes for Research Output) is not just citations — it's licensed, entitlement-checked content supply woven into the workspace. That's a moat destination-layer startups can't easily replicate, and it extends the free-incumbent pressure described in Scholar Labs: Google Scholar's Conservative "Deep Search" Entry — and the Academic AI Quadrant Map.
- A third publisher-payment model emerges. Parallel pays publishers per agent-use (Parallel: Deep Research Sold as Per-Request Backend Infrastructure — with Audit-Grade Provenance and Shapley-Value Payouts to Publishers); Sacra sells premium proprietary data as agent infrastructure (Sacra: Premium Private Market Research as Agent-Ready Infrastructure); Google converts reader purchases into AI-grounding entitlements — the user pays the publisher, Google takes the engagement. Publishers get a discovery channel instead of a payout.
- The roadmap line — "third-party subscriptions, business research reports, and textbooks" — points straight at premium research data. If Expert Intelligence grows beyond Play Books into subscriptions and business research reports, Google's grounding layer starts absorbing the exact supply that premium research-data vendors sell (see the registered watch). For Hey, Lefty, the takeaway is that model-agnostic orchestration (Destination vs. Backend: The Research-Tool Market Bifurcates — and Both Layers Converge on MCP) must also be source-agnostic: the durable workspace should ingest licensed, user-owned, and open sources alike rather than compete on content supply.