How Securly AI Transparency handles Google Gemini

Gemini now appears in many places across Google: its own app, Google Search, the Chrome browser, Workspace apps and Google Classroom. Securly AI Transparency can block or monitor some of these surfaces directly. Others run inside the browser or operating system where no filtering vendor can reach them. For those, Google's own Admin console settings and Chrome policies offer controls to disable them. This article sets out which is which, and how we recommend configuring both.

Before you start

The blocking and monitoring described here require the Securly Chrome extension on managed Chrome and Chromebook devices. Monitoring of AI prompts and responses also requires the Chrome Extension Integration setting in AI Transparency, under Policy Editor, Global settings.

On devices filtered through SmartPAC or DNS, Securly can block gemini.google.com and notebooklm.google.com at the domain level, but prompts and responses aren't captured.

Many Gemini features are restricted by Google to users aged 18 and over. Where a surface below is marked as 18+, students won't see it unless Google changes that policy, but the coverage we describe still applies to staff and to any accounts where you enable it.

Gemini app and NotebookLM

Gemini app (gemini.google.com): can be blocked by adding it to the block list under Policy Editor, Allow/Block. When it's allowed, students' prompts and Gemini's responses appear in the AI Transparency dashboard, and prompts that raise a safety concern are passed to Securly Aware if your district uses it. Prompts in the Gemini app that draw on Workspace or Classroom data are covered in the same way.

NotebookLM (notebooklm.google.com): can be blocked in the same way. When it's allowed, chat prompts and responses appear in the AI Transparency dashboard. Sources and files that students upload to a notebook aren't monitored.

Google Search

AI Mode: can be blocked with the Restrict AI-powered mode in search engines setting, under Policy Editor, then the policy's Settings. AI Mode queries appear in your regular Filter search logs.

AI Overviews: search queries are logged as normal. The Restrict AI-powered mode setting restricts AI Mode only; it doesn't currently hide AI Overviews in ordinary Google Search results.

Gemini in Chrome and Google Lens

The Gemini side panel in Chrome, the Google Lens overlay and Lens region search load outside normal web pages, so the Securly extension can't see into them. We recommend turning them off for students with these Chrome policies in the Google Admin console:

• Gemini side panel: GeminiIntegrationsEnabled

• Lens overlay: LensOverlaySettings

• Lens region search: LensRegionSearchEnabled

• AI Mode button in the Chrome toolbar: AIModeSettings (this removes the button only; AI Mode still appears in search results, so use the Securly setting above as well)

Gemini in Workspace apps (18+)

Gemini features in Gmail, Docs, Drive, Sheets, Slides, Forms, Vids, Meet, Chat and Calendar require a Gemini license and are restricted by Google to users aged 18 and over. If you enable Gemini in Google Docs for a group of users, prompts and responses in the Gemini side panel appear in the AI Transparency dashboard. Gemini features in Gmail, Drive, Meet, Chat and Calendar aren't monitored by Securly, so manage them in the Google Admin console under Generative AI.

Gemini in Google Classroom

The Gemini tab in Classroom, including Guided Learning and AI story generation in Read Along, is restricted by Google to users aged 18 and over.

Securly's blocking and monitoring are designed for Gemini and NotebookLM opened directly in the browser. Gems and notebooks that teachers share into Classroom, and Gemini embedded in other learning platforms such as Moodle, load inside another page, and coverage there may differ. We recommend managing access to these through the Google Admin console under Apps, Additional Google services, Gemini and NotebookLM.

Your data

Your district decides whether AI prompts and responses are captured, using the Chrome Extension Integration setting in AI Chat. Captured AI activity is kept for 15 months, in line with Filter activity. Prompts flagged by Securly Aware are kept for 48 months with other Aware alerts.

Recommended setup

1. Decide whether students should have access to the Gemini app and NotebookLM, and block or allow them in your Filter policies.

2. If you allow them, turn on Chrome Extension Integration in AI Chat so prompts and responses appear in the AI Transparency dashboard.

3. Turn on Restrict AI-powered mode in search engines for student policies if you don't want AI Mode available.

4. Apply the Chrome policies above to turn off the Gemini side panel and Lens for students.

5. Review Gemini settings in the Google Admin console under Generative AI for any staff or older-student groups you enable.

If your district approves an AI tool that doesn't appear in the AI Transparency dashboard, contact Securly Support so we can look at adding it.