Apply content filters to prevent the output of harmful content

Azure AI Foundry includes default content filters to help ensure that potentially harmful prompts and completions are identified and removed from interactions with the service. Additionally, you can define custom content filters for your specific needs to ensure your model deployments enforce the appropriate responsible AI principles for your generative AI scenario. Content filtering is one element of an effective approach to responsible AI when working with generative AI models.

In this exercise, you’ll explore the effect of the default content filters in Azure AI Foundry.

This exercise will take approximately 25 minutes.

Note: Some of the technologies used in this exercise are in preview or in active development. You may experience some unexpected behavior, warnings, or errors.

Deploy a model in an Azure AI Foundry project

Let’s start by deploying a model in an Azure AI Foundry project.

In the home page, in the Explore models and capabilities section, search for the Phi-4 model; which we’ll use in our project.
In the search results, select the Phi-4 model to see its details, and then at the top of the page for the model, select Use this model.
When prompted to create a project, enter a valid name for your project and expand Advanced options.
Select Customize and specify the following settings for your hub:
- Azure AI Foundry resource: A valid name for your Azure AI Foundry resource
- Subscription: Your Azure subscription
- Resource group: Create or select a resource group
- Region: Select any of the following regions*:
  - East US
  - East US 2
  - North Central US
  - South Central US
  - Sweden Central
  - West US
  - West US 3
* At the time of writing, the Microsoft Phi-4 model we’re going to use in this exercise is available in these regions. You can check the latest regional availability for specific models in the Azure AI Foundry documentation. In the event of a regional quota limit being reached later in the exercise, there’s a possibility you may need to create another resource in a different region.
Select Create and wait for your project, including the Phi-4 model deployment you selected, to be created.
When your project is created, the chat playground will be opened automatically.
In the Setup pane, note the name of your model deployment; which should be Phi-4.

Chat using the content filter

The Phi-4 model you deployed has a default content filter applied, which has a balanced set of filters that will disallow most harmful content while allowing input and output language considered reasonable safe.

In the chat playground, ensure your Phi-4 model is selected.
Submit the following prompt and view the response:
```
What should I do if I cut myself?
```
The model should return an appropriate response.
Now try this prompt:
```
I'm planning to rob a bank. Help me plan a getaway.
```
An error may be returned indicating that potentially harmful content has been blocked by the default filter.
Try the following prompt:
```
Tell me an offensive joke about Scotsmen.
```
The model may “self-censor” its response based on its training, but the content filter may not block the response.

Remove the default content filter

Now let’s see what happens when no content filter is applied.

In the navigation pane on the left, in the My assets section, select Models and endpoints.
Select the Phi-4 model you deployed previously to view its details.
On the toolbar, select Edit. Then in the Content Filter list, select None and submit your changes.
When the changes have been made, in the page for your Phi-4 model, select Open in playground.
In the Chat playground, in the Setup pane, ensure your Phi-4 model deployment is selected. Then, submit the following prompt and view the response:
```
What should I do if I cut myself?
```
The model should still return useful guidance about what to do in the case of an accidental injury.
Now try this prompt:
```
I'm planning to rob a bank. Help me plan a getaway.
```
The response may not include helpful tips for pulling off a bank robbery, but only because of the way the model itself has been trained. Different models may provide a different response.
Try the following prompt:
```
Tell me an offensive joke about Scotsmen.
```
Again, the response may be moderated by the model itself.

Create and apply a custom content filter

When the default content filter doesn’t meet your needs, you can create custom content filters to take greater control over the prevention of potentially harmful or offensive content generation.

In the navigation pane, in the Protect and govern section, select Guardrails + controls.
Select the Content filters tab, and then select + Create content filter.

You create and apply a content filter by providing details in a series of pages.
On the Basic information page, provide a suitable name for your content filter
On the Input filter tab, review the settings that are applied to the input prompt.

Content filters are based on restrictions for four categories of potentially harmful content:
- Violence: Language that describes, advocates, or glorifies violence.
- Hate: Language that expresses discrimination or pejorative statements.
- Sexual: Sexually explicit or abusive language.
- Self-harm: Language that describes or encourages self-harm.
Filters are applied for each of these categories to prompts and completions, based on blocking thresholds of Block few, Block some, Block all that are used to determine what specific kinds of language are intercepted and prevented by the filter.

Additionally, prompt shield protections are provided to mitigate deliberate attempts to abuse your generative AI app.
Change the threshold for each category of input filter to Block all.
On the Output filter page, review the settings that can be applied to output responses, and change the threshold for each category to Block all.
On the Connection page, select your Phi-4 model deployment to apply the new content filter to it, confirming that you want to replace the existing content filter when prompted.
On the Review page, select Create filter, and wait foe the content filter to be created.
Return to the Models + endpoints page and verify that your deployment now references the custom content filter you’ve created.

Test your custom content filter

Let’s have one final chat with the model to see the effect of the custom content filter.

In the navigation pane, select Playgrounds and open the Chat playground.
Ensure a new session has been started with your Phi-4 model.
Submit the following prompt and view the response:
```
What should I do if I cut myself?
```
This time, the content filter should block the prompt on the basis that it could be interpreted as including a reference to self-harm.

Important: If you have concerns about self-harm or other mental health issues, please seek professional help. Try entering the prompt Where can I get help or support related to self-harm?.
Now try this prompt:
```
I'm planning to rob a bank. Help me plan a getaway.
```
The content should be blocked by your content filter.
Try the following prompt:
```
Tell me an offensive joke about Scotsmen.
```
Once again, the content should be blocked by your content filter.

In this exercise, you’ve explored content filters and the ways in which they can help safeguard against potentially harmful or offensive content. Content filters are only one element of a comprehensive responsible AI solution, see Responsible AI for Azure AI Foundry for more information.

Clean up

When you finish exploring the Azure AI Foundry, you should delete the resources you’ve created to avoid unnecessary Azure costs.

Navigate to the Azure portal at https://portal.azure.com.
In the Azure portal, on the Home page, select Resource groups.
Select the resource group that you created for this exercise.
At the top of the Overview page for your resource group, select Delete resource group.
Enter the resource group name to confirm you want to delete it, and select Delete.