Google NotebookLM Rebrand May Expose Your Site To More AI Scraping

Search Engine Journal reports that Google’s rebrand of NotebookLM to Gemini Notebook comes with a user agent swap that has real operational consequences for anyone actively blocking this fetcher. The old Google-NotebookLM user agent string will stop working in August 2026, giving site owners a narrow window to update their blocking rules.

What’s actually new

The user agent identifier changed from Google-NotebookLM to Google-GeminiNotebook, with full mobile and desktop agent strings now documented by Google. The old identifier remains functional until August 2026 as a grace period. Functionally, Gemini Notebook behaves identically to NotebookLM — it’s a name change, not a product change. Notably, Project Mariner has been fully retired and removed from Google’s user-triggered fetcher documentation as well.

The critical detail here: Gemini Notebook’s fetcher is classified as a user-triggered fetcher, meaning it does not respect robots.txt. If you’ve been relying on robots.txt to block it, you were never blocking it in the first place. The only effective methods are firewall rules or .htaccess rewrites targeting the user agent string.

What it means for your config

This is a straightforward string replacement in your server config. If you have .htaccess rewrite rules or WAF/firewall rules matching Google-NotebookLM, you need to update them to match Google-GeminiNotebook before August 2026. The source provides a working .htaccess example using RewriteCond against %{HTTP_USER_AGENT} — check the original article for the exact snippet.

A few things worth thinking about on the config side:

• Nginx users: The article only shows an Apache .htaccess example. You’ll need to translate this to an if ($http_user_agent ...) block or handle it at the firewall level. The matching string is the same: Google-GeminiNotebook.
• Firewall-level blocking: If you’re using Cloudflare, AWS WAF, or similar, update any user agent rules referencing the old string. Consider matching both old and new strings during the transition period so you’re covered regardless of when Google flips the switch.
• robots.txt won’t help: This is worth repeating because it’s unintuitive. Google explicitly categorizes this as a user-triggered fetcher, so Disallow directives for this agent are simply ignored. Don’t waste config lines on it there.
• No interaction with other crawlers: This change is scoped to Gemini Notebook only. It doesn’t affect Googlebot, Google-Extended, or the Google-Agent fetcher (which is still documented separately for agent-based browsing).

Recommended next step

Grep your server configs, .htaccess files, and firewall rules for Google-NotebookLM right now. If you find matches, add rules for Google-GeminiNotebook alongside the existing ones — keep both active until the old agent is fully retired in August 2026, then clean up. If you haven’t been blocking this fetcher and you care about your content being scraped for AI-generated summaries, audio, and video without attribution or referral traffic, this is a good time to decide whether you want to start.

---

Read the full announcement on Search Engine Journal → Google NotebookLM Rebrand May Expose Your Site To More AI Scraping