GSP1327

Overview

Model Armor on Google Cloud enhances AI safety by screening both user prompts and LLM responses for risks such as harmful content, prompt injection, and sensitive data leaks. It allows organizations to centrally manage and enforce security policies for their AI applications, acting as a filter to ensure interactions are safe and compliant. This service integrates with Security Command Center for unified visibility and helps protect against unauthorized exposure or manipulation of AI models.

In this lab, you explore sanitizing prompts and their responses using Model Armor.

Objectives

This lab challenges you to demonstrate your ability to do the following:

• Enable the Model Armor API.
• Create a Model Armor template.
• Configure Sensitive Data Protection with Inspect and De-identify templates.
• Execute various commands for sanitizing user prompts against different security features.

Setup and requirements

Before you click the Start Lab button

Read these instructions. Labs are timed and you cannot pause them. The timer, which starts when you click Start Lab, shows how long Google Cloud resources are made available to you.

This hands-on lab lets you do the lab activities in a real cloud environment, not in a simulation or demo environment. It does so by giving you new, temporary credentials you use to sign in and access Google Cloud for the duration of the lab.

To complete this lab, you need:

• Access to a standard internet browser (Chrome browser recommended).

• Time to complete the lab—remember, once you start, you cannot pause a lab.

How to start your lab and sign in to the Google Cloud console

1. Click the Start Lab button. If you need to pay for the lab, a dialog opens for you to select your payment method. On the right is the Lab setup and access panel with the following:

   • The Open Google Cloud console button
   • The temporary credentials (username and password) that you must use for this lab
   • Other information, if needed, to step through this lab

   Note that the lab timer is located near the top of the page, showing the remaining time.

2. Click Open Google Cloud console (or right-click and select Open Link in Incognito Window if you are running the Chrome browser).

   The lab spins up resources, and then opens another tab that shows the Sign in page.

   Tip: Arrange the tabs in separate windows, side-by-side.

   Note: If you see the Choose an account dialog, click Use Another Account.

3. If necessary, copy the Username below and paste it into the Sign in dialog.

   {{{user_0.username | "Username"}}}

   You can also find the Username in the Lab setup and access panel.

4. Click Next.

5. Copy the Password below and paste it into the Welcome dialog.

   {{{user_0.password | "Password"}}}

   You can also find the Password in the Lab setup and access panel.

6. Click Next.

   Important: You must use the credentials the lab provides you. Do not use your Google Cloud account credentials. Note: Using your own Google Cloud account for this lab may incur extra charges.

7. Click through the subsequent pages:

   • Accept the terms and conditions.
   • Do not add recovery options or two-factor authentication (because this is a temporary account).
   • Do not sign up for free trials.

After a few moments, the Google Cloud console opens in this tab.

Note: To access Google Cloud products and services, click the Navigation menu or type the service or product name in the Search field.

Task 1. Enable the Model Armor API

Before you can begin using Model Armor, you must enable the API.

1. In the Google Cloud console, activate Cloud Shell.

2. Execute the following command:

Click Check my progress to verify the objective. Enable the Model Armor API

Task 2. Open the notebook in Agent Platform Workbench

1. In the Google Cloud console, on the Navigation menu (), click Agent Platform > Notebooks > Workbench.

2. Find the instance and click on the Open JupyterLab button.

The JupyterLab interface for your Workbench instance opens in a new browser tab.

Task 3. Set up the notebook

1. Open the file.

2. In the Select Kernel dialog, choose Python 3 from the list of available kernels.

3. Run through the Getting Started and the Import libraries sections of the notebook.

Task 4. Create a Model Armor template

In this task, you create a Model Armor template, ma-template, to define security and content filtering rules for AI interactions. This template establishes policies for detecting and mitigating sensitive data exposure, ensuring compliance with data protection standards.

Run through the Create a Model Armor Template section of the notebook to create the ma-template in the location using the provided filter configuration. Once completed, return here to check your progress and verify the objective.

Click Check my progress to verify the objective. Create a Model Armor template

Task 5. Configure Sensitive Data Protection with Inspect and De-identify templates

In this task, you create Cloud Sensitive Data Protection (DLP) Inspect and De-identify templates, and link them to an advanced Model Armor template, ma-sdp-template, using advancedConfig. This enables custom inspection rules and automated data masking for AI prompt and response sanitization.

1. Run through the Configure Sensitive Data Protection with Inspect and De-identify Templates section of the notebook to create ma-inspect-template and ma-deidentify-template.

2. Execute the cells to create the advanced Model Armor template ma-sdp-template using advancedConfig.

3. Test sanitizing model responses to observe sensitive infoTypes being de-identified.

Click Check my progress to verify the objective. Configure Sensitive Data Protection with Inspect and De-identify templates

Task 6. Validate user prompts with Model Armor

In this task, you test various user prompts to evaluate how Model Armor detects and screens different types of sensitive data based on predefined Sensitive Data Protection (SDP) infoTypes. This process helps ensure that Model Armor effectively filters and protects against unauthorized data exposure.

Go through the Validate user prompts with Model Armor section of the notebook to execute various commands for sanitizing user prompts against different security features. Once complete, return here to check your progress and verify the objectives.

1. Input a user prompt for Model Armor to screen Responsible AI.

Click Check my progress to verify the objective. Input a user prompt to screen Responsible AI

2. Input a user prompt for Model Armor to screen Malicious URI.

Click Check my progress to verify the objective. Input a user prompt to screen Malicious URI

3. Input a user prompt for Model Armor to screen for data loss prevention (DLP).

Click Check my progress to verify the objective. Input a user prompt to screen DLP

4. Input a model response for Model Armor to screen DLP.

Click Check my progress to verify the objective. Input a model response to screen DLP

5. Run the cell in the Sanitize file-based prompts subsection to execute a file-based prompt that uses Model Armor to sanitize a user prompt in the specified file format.

Click Check my progress to verify the objective. Sanitize a file-based user prompt

Task 7. View Model Armor logs in Logs Explorer

In this task, you inspect the Model Armor sanitization operation logs in Cloud Logging to review prompt and response screening details, filter matches, and evaluation metadata.

1. In the Google Cloud console navigation menu, click Monitoring > Logs Explorer.

2. In the Query pane, enter the following log query:

3. Ensure the time range filter (top-right next to Run Query) is set to Last 1 hour.

4. Click Run Query.

5. Inspect the returned log entries to review the Model Armor sanitization results and filter evaluation details.

Congratulations!

You have successfully completed the lab to create Model Armor templates, configure Sensitive Data Protection Inspect and De-identify templates, execute prompts and model responses to screen and de-identify sensitive data, and view sanitization logs in Logs Explorer.

Manual Last Updated August 24, 2026

Lab Last Tested August 24, 2026

Copyright 2026 Google LLC. All rights reserved. Google and the Google logo are trademarks of Google LLC. All other company and product names may be trademarks of the respective companies with which they are associated.