NotebookLM now has the ability to automatically arrange your research sources.
Managing sources in NotebookLM has just become a breeze.
If you're utilizing NotebookLM for research, you're likely aware of how quickly sources can accumulate. Handling them manually, particularly in notebooks with ten or more entries, has been one of the most frustrating challenges of the tool, but Google has resolved this issue.
NotebookLM, the AI-driven research assistant powered by Gemini, is introducing automatic source labeling and categorization. This feature activates once you have five or more sources in a notebook, and it automatically assigns labels for you.
Got more sources, more problems? Not anymore: Currently being rolled out, NotebookLM can automatically label and categorize sources (when you have five or more), allowing you to spend less time scrolling and more time thinking, learning, and philosophizing. You can rename, reorganize, and personalize (even with emojis!) according to your preferences. pic.twitter.com/WiY58zkQJU— NotebookLM (@NotebookLM) April 24, 2026
How does NotebookLM's auto-labeling and source categorization feature function?
When your notebook exceeds the five-source limit, NotebookLM analyzes each source's content and groups related ones together. It then assigns labels to these groups based on their subject matter. If a single source covers multiple topics, NotebookLM can assign it more than one label, ensuring your organization remains flexible and precise.
The feature aims to reduce the time spent scrolling through an unorganized collection of materials to find what you truly need. You maintain complete control over the results; labels can be renamed, rearranged, and even customized with emojis. If you disagree with how NotebookLM has categorized something, you can override it and apply your own label.
Google is also contemplating further enhancements to this feature to improve the organization of outputs, though that upgrade has yet to be confirmed.
What other new features are on their way to NotebookLM?
Notebook sharing is also getting a beneficial update. Previously, sharing a notebook with a group required entering each email address individually, which was cumbersome for larger teams.
Another minor update for NotebookLM: ✅ Sharing notebooks with multiple people at once has just become much simpler. No more entering emails one by one—just copy and paste your list, hit send, and get ready for the world to see your creation. pic.twitter.com/RyjXtHcqtW— NotebookLM (@NotebookLM) April 23, 2026
You can now paste an entire list of email addresses at once, and NotebookLM will automatically parse and identify the recipients. Both features are currently rolling out and should soon be available to all users of NotebookLM.
As Google continues to strengthen the integration of NotebookLM and Gemini, the tool has recently been incorporated into Gemini Notebooks, and Notebook projects are now free for all Gemini users on the web, providing more reasons to incorporate it into daily workflows.
Manisha is a Writer at Digital Trends, covering the latest in tech, science, AI, gaming, and entertainment. As a Computer…
Samsung’s new Snapdragon-powered Galaxy Book isn’t the MacBook Neo rival I was hoping for
Galaxy Book 6 Edge leaks with ARM power and hefty pricing.
The latest Snapdragon-driven laptop from Samsung has recently leaked in full, showcasing some premium specs and ARM capabilities. However, it doesn’t appear to be the MacBook Neo-style competitor that I had hoped for. Newly leaked renders and listings for the Galaxy Book 6 Edge give us a near-complete understanding of what Samsung is working on, although the overall direction feels quite familiar.
Why this isn't a new entry-level Samsung laptop
Microsoft revamps Windows Insider Program with simpler structure and more user control
I’m pleased that Microsoft has streamlined the Insider program—it was long overdue.
Microsoft is launching a significant revamp of its Windows Insider Program, intending to simplify the process through which early Windows features are tested while also addressing longstanding user concerns about complexity and control. This update represents one of the most substantial structural changes to the program in years, indicating a new direction in how Microsoft aims to collaborate with its testing community.
A Simpler Insider Program Built Around Feedback And Control
Discord users breach access controls to reach Anthropic’s Mythos model
This AI security breach underscores the importance of protecting your data.
A recent security incident involving Anthropic has revealed the vulnerability of safeguards surrounding advanced AI systems. A Wired report indicates that a small group of users, operating via private Discord channels, managed to gain unauthorized access to the company's highly restricted Mythos AI model—an experimental system designed for cybersecurity applications.
A Breach That Exposes Bigger Risks Around AI Control
Other articles
China establishes protections for gig workers, impacting 200 million platform employees, with an emphasis on algorithm transparency and a deadline set for 2027.
China's top governing authorities have required minimum wage, maximum working hours, and transparency in algorithms for over 200 million gig workers. Applications are prohibited from dispatching orders once drivers reach their time limits. Compliance is expected by 2027.
Discord users bypass access restrictions to gain entry to Anthropic's Mythos model.
Unauthorized access to Anthropic’s Mythos AI underscores increasing worries about protecting powerful systems, revealing weaknesses not within the model itself but in the access framework that surrounds it.
The trial between Musk and Altman commences, with $150 billion on the line regarding OpenAI's shift from nonprofit to for-profit status.
Jury selection will begin on Monday for the Musk-Altman trial. In his 2017 diary, Brockman referred to OpenAI's commitment to being a nonprofit as "a lie." The judge has found "ample evidence" and will independently determine the remedies.
Sam Altman has issued an apology following OpenAI's decision not to report a ChatGPT user involved in the Tumbler Ridge school shooting.
Eight months prior to the attack, OpenAI marked the ChatGPT account of the Tumbler Ridge shooter as a concern. Staff suggested notifying law enforcement, but their recommendations were dismissed by management. Altman's apology cannot bridge this gap.
I never imagined that AI would introduce typos – but it actually makes some sense.
A new AI tool intentionally introduces typos into emails, highlighting a trend where flawed writing is perceived as more genuine and reliable in an era of AI-generated messages.
Sequoia hands out 200 engraved Mac Minis at the AI event as OpenClaw establishes itself as the infrastructure layer that venture capitalists cannot possess.
Alfred Lin distributed 200 engraved Mac Minis that operate on OpenClaw, the open-source AI agent framework that has outperformed Apple's stock and exceeded React on GitHub. Sequoia is unable to invest in it. That is the approach being used.
NotebookLM now has the ability to automatically arrange your research sources.
Google has introduced automatic source labeling and categorization for NotebookLM. This feature activates automatically when you have five or more sources in a notebook.