Google upgrades NotebookLM with Gemini 3.5 and source-building chat feature
The AMW Read
Incremental update to a known Google product; no structural shift or new entrant.
Google upgrades NotebookLM with Gemini 3.5 and source-building chat feature
Google announced an update to NotebookLM, its AI-powered research note-taking tool, migrating it to the Gemini 3.5 model as the default and introducing a feature that lets users build a source repository directly through chat. The new capability allows NotebookLM to suggest sources via Google Search and its research skills, moving beyond the previous requirement that users upload all source material manually. The update also adds output formatting options, exports to multiple file types (PDF, DOCX, CSV, JSON, etc.), and detailed reasoning steps for answers. Availability begins today for Google AI Ultra and Workspace business customers with AI Ultra or AI Expanded Access.
Why it matters: This update tightens Google’s grip on the enterprise research workflow by embedding its foundation model deeper into a product that previously required manual curation. By shifting from a “bring your own sources” model to a proactive research assistant that builds knowledge bases, NotebookLM now directly competes with purpose-built research tools and agents from startups like Perplexity and Glean. The Gemni 3.5 upgrade and step-by-step reasoning align with the broader industry move toward transparent, verifiable AI outputs—a key trust requirement for enterprise adoption. This is also a clear example of a hyperscaler leveraging its search and infrastructure moat to fold adjacent use cases into its ecosystem, making it harder for standalone players to differentiate purely on research quality.
Grounded expert take: Google’s play here mirrors the “hyperscaler-distribution” pattern where the owner of both search and a foundation model can rapidly embed generative capabilities into existing tools with massive user bases. NotebookLM’s evolution from a passive document analyzer to an active research agent—complete with source suggestions and multilingual material discovery—signals a strategic expansion into the professional knowledge-management vertical. The migration to Gemini 3.5 also suggests Google trusts its in-house model enough to make it default, a vote of confidence that could influence enterprise buyer decisions. However, the paid-tier exclusivity (AI Ultra/Workspace) indicates this is a monetization lever rather than a pure consumer play, and the lack of pricing details leaves open questions about adoption velocity.
