Manage Azure AI Services Security

Security is a critical consideration for any application, and as a developer you should ensure that access to resources such as Azure AI services is restricted to only those who require it.

Access to Azure AI services is typically controlled through authentication keys, which are generated when you initially create an Azure AI services resource.

Clone the repository for this course

If you have already cloned AI-102-AIEngineer code repository to the environment where you’re working on this lab, open it in Visual Studio Code; otherwise, follow these steps to clone it now.

  1. Start Visual Studio Code.
  2. Open the palette (SHIFT+CTRL+P) and run a Git: Clone command to clone the https://github.com/MicrosoftLearning/AI-102-AIEngineer repository to a local folder (it doesn’t matter which folder).
  3. When the repository has been cloned, open the folder in Visual Studio Code.

  4. Wait while additional files are installed to support the C# code projects in the repo.

    Note: If you are prompted to add required assets to build and debug or update Visual Studio Code, select Not Now.

Provision an Azure AI Services resource

If you don’t already have one in your subscription, you’ll need to provision an Azure AI Services resource.

  1. Open the Azure portal at https://portal.azure.com, and sign in using the Microsoft account associated with your Azure subscription.
  2. In the top search bar, search for Azure AI services, select Azure AI Services, and create an Azure AI services multi-service account resource with the following settings:
    • Subscription: Your Azure subscription
    • Resource group: Choose or create a resource group (if you are using a restricted subscription, you may not have permission to create a new resource group - use the one provided)
    • Region: Choose any available region
    • Name: Enter a unique name
    • Pricing tier: Standard S0
  3. Select the required checkboxes and create the resource.
  4. Wait for deployment to complete, and then view the deployment details.

Manage authentication keys

When you created your Azure AI services resource, two authentication keys were generated. You can manage these in the Azure portal or by using the Azure command line interface (CLI).

  1. In the Azure portal, go to your Azure AI services resource and view its Keys and Endpoint page. This page contains the information that you will need to connect to your resource and use it from applications you develop. Specifically:
    • An HTTP endpoint to which client applications can send requests.
    • Two keys that can be used for authentication (client applications can use either of the keys. A common practice is to use one for development, and another for production. You can easily regenerate the development key after developers have finished their work to prevent continued access).
    • The location where the resource is hosted. This is required for requests to some (but not all) APIs.

  2. In Visual Studio Code, right-click the 02-cognitive-security folder and open an integrated terminal. Then enter the following command to sign into your Azure subscription by using the Azure CLI.

     az login

    A web browser tab will open and prompt you to sign into Azure. Do so, and then close the browser tab and return to Visual Studio Code.

    Tip: If you have multiple subscriptions, you’ll need to ensure that you are working in the one that contains your Azure AI services resource. Use the command az account show to determine your current subscription - its unique ID is the id value in the JSON that gets returned.

    Warning: If you are getting a certificate verfication failure for az login, try waiting a few minutes and trying again. If you need to change the subscription, run this command, changing <Your_Subscription_Id> to the correct subscription ID: az account set --subscription <Your_Subscription_Id> Alternatively, you can explicitly specify the subscription ID as a –subscription parameter in each Azure CLI command that follows.

  3. Now you can use the following command to get the list of Azure AI services keys, replacing <resourceName> with the name of your Azure AI services resource, and <resourceGroup> with the name of the resource group in which you created it.

     az cognitiveservices account keys list --name <resourceName> --resource-group <resourceGroup>

The command returns a list of the keys for your Azure AI services resource - there are two keys, named key1 and key2.

  1. To test your Azure AI service, you can use curl - a command line tool for HTTP requests. In the 02-cognitive-security folder, open rest-test.cmd and edit the curl command it contains (shown below), replacing <yourEndpoint> and <yourKey> with your endpoint URI and Key1 key to use the Text Analytics API in your Azure AI services resource.

     curl -X POST "<yourEndpoint>/text/analytics/v3.0/languages?" -H "Content-Type: application/json" -H "Ocp-Apim-Subscription-Key: <yourKey>" --data-ascii "{'documents':[{'id':1,'text':'hello'}]}"

  2. Save your changes, and then in the integrated terminal for the 02-cognitive-security folder, run the following command:

     rest-test

The command returns a JSON document containing information about the language detected in the input data (which should be English).

  1. If a key becomes compromised, or the developers who have it no longer require access, you can regenerate it in the portal or by using the Azure CLI. Run the following command to regenerate your key1 key (replacing <resourceName> and <resourceGroup> for your resource).

     az cognitiveservices account keys regenerate --name <resourceName> --resource-group <resourceGroup> --key-name key1

The list of keys for your Azure AI services resource is returned - note that key1 has changed since you last retrieved them.

  1. Re-run the rest-test command with the old key (you can use the ^ key to cycle through previous commands), and verify that it now fails.
  2. Edit the curl command in rest-test.cmd replacing the key with the new key1 value, and save the changes. Then rerun the rest-test command and verify that it succeeds.

Tip: In this exercise, you used the full names of Azure CLI parameters, such as –resource-group. You can also use shorter alternatives, such as -g, to make your commands less verbose (but a little harder to understand). The Azure AI Services CLI command reference lists the parameter options for each Azure AI services CLI command.

Secure key access with Azure Key Vault

You can develop applications that consume Azure AI services by using a key for authentication. However, this means that the application code must be able to obtain the key. One option is to store the key in an environment variable or a configuration file where the application is deployed, but this approach leaves the key vulnerable to unauthorized access. A better approach when developing applications on Azure is to store the key securely in Azure Key Vault, and provide access to the key through a managed identity (in other words, a user account used by the application itself).

Create a key vault and add a secret

First, you need to create a key vault and add a secret for the Azure AI services key.

  1. Make a note of the key1 value for your Azure AI services resource (or copy it to the clipboard).

  2. In the Azure portal, on the Home page, select the ＋Create a resource button, search for Key Vault, and create a Key Vault resource with the following settings:

    • Basics tab
      • Subscription: Your Azure subscription
      • Resource group: The same resource group as your Azure AI service resource
      • Key vault name: Enter a unique name
      • Region: The same region as your Azure AI service resource
      • Pricing tier: Standard
    • Access configuration tab
      • Permission model: Vault access policy
      • Scroll down to the Access policies section and select your user using the checkbox on the left.
  3. Wait for deployment to complete and then go to your key vault resource.
  4. In the left navigation pane, select Secrets (in the Objects section).
  5. Select + Generate/Import and add a new secret with the following settings :
    • Upload options: Manual
    • Name: Cognitive-Services-Key (it’s important to match this exactly, because later you’ll run code that retrieves the secret based on this name)
    • Value: Your key1 Azure AI services key

Create a service principal

To access the secret in the key vault, your application must use a service principal that has access to the secret. You’ll use the Azure command line interface (CLI) to create the service principal, find its object ID, and grant access to the secret in Azure Vault.

  1. Return to Visual Studio Code, and in the integrated terminal for the 02-cognitive-security folder, run the following Azure CLI command, replacing <spName> with a unique suitable name for an application identity (for example, ai-app with your initials appended on the end; the name must be unique within your tenant). Also replace <subscriptionId> and <resourceGroup> with the correct values for your subscription ID and the resource group containing your Azure AI services and key vault resources:

    Tip: If you are unsure of your subscription ID, use the az account show command to retrieve your subscription information - the subscription ID is the id attribute in the output. If you see an error about the object already existing, please choose a different unique name.

     az ad sp create-for-rbac -n "api://<spName>" --role owner --scopes subscriptions/<subscriptionId>/resourceGroups/<resourceGroup>

The output of this command includes information about your new service principal. It should look similar to this:

```
{
    "appId": "abcd12345efghi67890jklmn",
    "displayName": "ai-app",
    "name": "http://ai-app",
    "password": "1a2b3c4d5e6f7g8h9i0j",
    "tenant": "1234abcd5678fghi90jklm"
}
```

Make a note of the appId, password, and tenant values - you will need them later (if you close this terminal, you won’t be able to retrieve the password; so it’s important to note the values now - you can paste the output into a new text file in Visual Studio Code to ensure you can find the values you need later!)

  1. To get the object ID of your service principal, run the following Azure CLI command, replacing <appId> with the value of your service principal’s app ID. If running the following command gives no response, you may be using a different version of Azure CLI; replace objectId with id if that is the case.

     az ad sp show --id <appId> --query objectId --out tsv

  2. To assign permission for your new service principal to access secrets in your Key Vault, run the following Azure CLI command, replacing <keyVaultName> with the name of your Azure Key Vault resource and <objectId> with the value of your service principal’s object ID.

     az keyvault set-policy -n <keyVaultName> --object-id <objectId> --secret-permissions get list

Use the service principal in an application

Now you’re ready to use the service principal identity in an application, so it can access the secret Azure AI services key in your key vault and use it to connect to your Azure AI services resource.

Note: In this exercise, we’ll store the service principal credentials in the application configuration and use them to authenticate a ClientSecretCredential identity in your application code. This is fine for development and testing, but in a real production application, an administrator would assign a managed identity to the application so that it uses the service principal identity to access resources, without caching or storing the password.

  1. In Visual Studio Code, expand the 02-cognitive-security folder and the C-Sharp or Python folder depending on your language preference.

  2. Right-click the keyvault-client folder and open an integrated terminal. Then install the packages you will need to use Azure Key Vault and the Text Analytics API in your Azure AI services resource by running the appropriate command for your language preference:

    C#

     dotnet add package Azure.AI.TextAnalytics --version 5.3.0
 dotnet add package Azure.Identity --version 1.5.0
 dotnet add package Azure.Security.KeyVault.Secrets --version 4.2.0-beta.3

    Python

     pip install azure-ai-textanalytics==5.3.0
 pip install azure-identity==1.5.0
 pip install azure-keyvault-secrets==4.2.0
  3. View the contents of the keyvault-client folder, and note that it contains a file for configuration settings:
    • C#: appsettings.json
    • Python: .env

    Open the configuration file and update the configuration values it contains to reflect the following settings:

    • The endpoint for your Azure AI Services resource
    • The name of your Azure Key Vault resource
    • The tenant for your service principal
    • The appId for your service principal
    • The password for your service principal

    Save your changes.

  4. Note that the keyvault-client folder contains a code file for the client application:

    • C#: Program.cs
    • Python: keyvault-client.py

    Open the code file and review the code it contains, noting the following details:

    • The namespace for the SDK you installed is imported
    • Code in the Main function retrieves the application configuration settings, and then it uses the service principal credentials to get the Azure AI services key from the key vault.
    • The GetLanguage function uses the SDK to create a client for the service, and then uses the client to detect the language of the text that was entered.

  5. Return to the integrated terminal for the keyvault-client folder, and enter the following command to run the program:

    C#

     dotnet run

    Python

     python keyvault-client.py
  6. When prompted, enter some text and review the language that is detected by the service. For example, try entering “Hello”, “Bonjour”, and “Gracias”.
  7. When you have finished testing the application, enter “quit” to stop the program.

More information

For more information about securing Azure AI services, see the Azure AI Services security documentation.