Rai Policies - Create Or Update

Service:: Azure AI Services

API Version:: 2024-10-01

Update the state of specified Content Filters associated with the Azure OpenAI account.

PUT https://management.azure.com/subscriptions/{subscriptionId}/resourceGroups/{resourceGroupName}/providers/Microsoft.CognitiveServices/accounts/{accountName}/raiPolicies/{raiPolicyName}?api-version=2024-10-01

URI Parameters

Name	In	Required	Type	Description
accountName	path	True	string minLength: 2 maxLength: 64 pattern: ^[a-zA-Z0-9][a-zA-Z0-9_.-]*$	The name of Cognitive Services account.
raiPolicyName	path	True	string pattern: ^[a-zA-Z0-9][a-zA-Z0-9_.-]*$	The name of the RaiPolicy associated with the Cognitive Services Account
resourceGroupName	path	True	string minLength: 1 maxLength: 90	The name of the resource group. The name is case insensitive.
subscriptionId	path	True	string minLength: 1	The ID of the target subscription.
api-version	query	True	string minLength: 1	The API version to use for this operation.

Request Body

Name	Type	Description
properties	RaiPolicyProperties	Properties of Cognitive Services RaiPolicy.
tags	object	Resource tags.

Responses

Name	Type	Description
200 OK	RaiPolicy	Create/Update the Content Filters successfully.
201 Created	RaiPolicy	Create the Content Filters successfully.
Other Status Codes	ErrorResponse	Error response describing why the operation failed.

Examples

PutRaiPolicy

Sample request

PUT https://management.azure.com/subscriptions/00000000-0000-0000-0000-000000000000/resourceGroups/resourceGroupName/providers/Microsoft.CognitiveServices/accounts/accountName/raiPolicies/raiPolicyName?api-version=2024-10-01

{
  "properties": {
    "basePolicyName": "Microsoft.Default",
    "mode": "Asynchronous_filter",
    "contentFilters": [
      {
        "name": "Hate",
        "blocking": false,
        "enabled": false,
        "severityThreshold": "High",
        "source": "Prompt"
      },
      {
        "name": "Hate",
        "blocking": true,
        "enabled": true,
        "severityThreshold": "Medium",
        "source": "Completion"
      },
      {
        "name": "Sexual",
        "blocking": true,
        "enabled": true,
        "severityThreshold": "High",
        "source": "Prompt"
      },
      {
        "name": "Sexual",
        "blocking": true,
        "enabled": true,
        "severityThreshold": "Medium",
        "source": "Completion"
      },
      {
        "name": "Selfharm",
        "blocking": true,
        "enabled": true,
        "severityThreshold": "High",
        "source": "Prompt"
      },
      {
        "name": "Selfharm",
        "blocking": true,
        "enabled": true,
        "severityThreshold": "Medium",
        "source": "Completion"
      },
      {
        "name": "Violence",
        "blocking": true,
        "enabled": true,
        "severityThreshold": "Medium",
        "source": "Prompt"
      },
      {
        "name": "Violence",
        "blocking": true,
        "enabled": true,
        "severityThreshold": "Medium",
        "source": "Completion"
      },
      {
        "name": "Jailbreak",
        "blocking": true,
        "source": "Prompt",
        "enabled": true
      },
      {
        "name": "Protected Material Text",
        "blocking": true,
        "source": "Completion",
        "enabled": true
      },
      {
        "name": "Protected Material Code",
        "blocking": true,
        "source": "Completion",
        "enabled": true
      },
      {
        "name": "Profanity",
        "blocking": true,
        "source": "Prompt",
        "enabled": true
      }
    ]
  }
}


import com.azure.resourcemanager.cognitiveservices.models.ContentLevel;
import com.azure.resourcemanager.cognitiveservices.models.RaiPolicyContentFilter;
import com.azure.resourcemanager.cognitiveservices.models.RaiPolicyContentSource;
import com.azure.resourcemanager.cognitiveservices.models.RaiPolicyMode;
import com.azure.resourcemanager.cognitiveservices.models.RaiPolicyProperties;
import java.util.Arrays;

/**
 * Samples for RaiPolicies CreateOrUpdate.
 */
public final class Main {
    /*
     * x-ms-original-file:
     * specification/cognitiveservices/resource-manager/Microsoft.CognitiveServices/stable/2024-10-01/examples/
     * PutRaiPolicy.json
     */
    /**
     * Sample code: PutRaiPolicy.
     * 
     * @param manager Entry point to CognitiveServicesManager.
     */
    public static void putRaiPolicy(com.azure.resourcemanager.cognitiveservices.CognitiveServicesManager manager) {
        manager.raiPolicies().define("raiPolicyName").withExistingAccount("resourceGroupName", "accountName")
            .withProperties(new RaiPolicyProperties().withMode(RaiPolicyMode.ASYNCHRONOUS_FILTER)
                .withBasePolicyName("Microsoft.Default")
                .withContentFilters(Arrays.asList(
                    new RaiPolicyContentFilter().withName("Hate").withEnabled(false)
                        .withSeverityThreshold(ContentLevel.HIGH).withBlocking(false)
                        .withSource(RaiPolicyContentSource.PROMPT),
                    new RaiPolicyContentFilter().withName("Hate").withEnabled(true)
                        .withSeverityThreshold(ContentLevel.MEDIUM).withBlocking(true)
                        .withSource(RaiPolicyContentSource.COMPLETION),
                    new RaiPolicyContentFilter().withName("Sexual").withEnabled(true)
                        .withSeverityThreshold(ContentLevel.HIGH).withBlocking(true)
                        .withSource(RaiPolicyContentSource.PROMPT),
                    new RaiPolicyContentFilter().withName("Sexual").withEnabled(true)
                        .withSeverityThreshold(ContentLevel.MEDIUM).withBlocking(true)
                        .withSource(RaiPolicyContentSource.COMPLETION),
                    new RaiPolicyContentFilter().withName("Selfharm").withEnabled(true)
                        .withSeverityThreshold(ContentLevel.HIGH).withBlocking(true)
                        .withSource(RaiPolicyContentSource.PROMPT),
                    new RaiPolicyContentFilter().withName("Selfharm").withEnabled(true)
                        .withSeverityThreshold(ContentLevel.MEDIUM).withBlocking(true)
                        .withSource(RaiPolicyContentSource.COMPLETION),
                    new RaiPolicyContentFilter().withName("Violence").withEnabled(true)
                        .withSeverityThreshold(ContentLevel.MEDIUM).withBlocking(true)
                        .withSource(RaiPolicyContentSource.PROMPT),
                    new RaiPolicyContentFilter().withName("Violence").withEnabled(true)
                        .withSeverityThreshold(ContentLevel.MEDIUM).withBlocking(true)
                        .withSource(RaiPolicyContentSource.COMPLETION),
                    new RaiPolicyContentFilter().withName("Jailbreak").withEnabled(true).withBlocking(true)
                        .withSource(RaiPolicyContentSource.PROMPT),
                    new RaiPolicyContentFilter().withName("Protected Material Text").withEnabled(true)
                        .withBlocking(true).withSource(RaiPolicyContentSource.COMPLETION),
                    new RaiPolicyContentFilter().withName("Protected Material Code").withEnabled(true)
                        .withBlocking(true).withSource(RaiPolicyContentSource.COMPLETION),
                    new RaiPolicyContentFilter().withName("Profanity").withEnabled(true).withBlocking(true)
                        .withSource(RaiPolicyContentSource.PROMPT))))
            .create();
    }
}

To use the Azure SDK library in your project, see this documentation. To provide feedback on this code sample, open a GitHub issue

from azure.identity import DefaultAzureCredential

from azure.mgmt.cognitiveservices import CognitiveServicesManagementClient

"""
# PREREQUISITES
    pip install azure-identity
    pip install azure-mgmt-cognitiveservices
# USAGE
    python put_rai_policy.py

    Before run the sample, please set the values of the client ID, tenant ID and client secret
    of the AAD application as environment variables: AZURE_CLIENT_ID, AZURE_TENANT_ID,
    AZURE_CLIENT_SECRET. For more info about how to get the value, please see:
    https://docs.microsoft.com/azure/active-directory/develop/howto-create-service-principal-portal
"""


def main():
    client = CognitiveServicesManagementClient(
        credential=DefaultAzureCredential(),
        subscription_id="00000000-0000-0000-0000-000000000000",
    )

    response = client.rai_policies.create_or_update(
        resource_group_name="resourceGroupName",
        account_name="accountName",
        rai_policy_name="raiPolicyName",
        rai_policy={
            "properties": {
                "basePolicyName": "Microsoft.Default",
                "contentFilters": [
                    {
                        "blocking": False,
                        "enabled": False,
                        "name": "Hate",
                        "severityThreshold": "High",
                        "source": "Prompt",
                    },
                    {
                        "blocking": True,
                        "enabled": True,
                        "name": "Hate",
                        "severityThreshold": "Medium",
                        "source": "Completion",
                    },
                    {
                        "blocking": True,
                        "enabled": True,
                        "name": "Sexual",
                        "severityThreshold": "High",
                        "source": "Prompt",
                    },
                    {
                        "blocking": True,
                        "enabled": True,
                        "name": "Sexual",
                        "severityThreshold": "Medium",
                        "source": "Completion",
                    },
                    {
                        "blocking": True,
                        "enabled": True,
                        "name": "Selfharm",
                        "severityThreshold": "High",
                        "source": "Prompt",
                    },
                    {
                        "blocking": True,
                        "enabled": True,
                        "name": "Selfharm",
                        "severityThreshold": "Medium",
                        "source": "Completion",
                    },
                    {
                        "blocking": True,
                        "enabled": True,
                        "name": "Violence",
                        "severityThreshold": "Medium",
                        "source": "Prompt",
                    },
                    {
                        "blocking": True,
                        "enabled": True,
                        "name": "Violence",
                        "severityThreshold": "Medium",
                        "source": "Completion",
                    },
                    {"blocking": True, "enabled": True, "name": "Jailbreak", "source": "Prompt"},
                    {"blocking": True, "enabled": True, "name": "Protected Material Text", "source": "Completion"},
                    {"blocking": True, "enabled": True, "name": "Protected Material Code", "source": "Completion"},
                    {"blocking": True, "enabled": True, "name": "Profanity", "source": "Prompt"},
                ],
                "mode": "Asynchronous_filter",
            }
        },
    )
    print(response)


# x-ms-original-file: specification/cognitiveservices/resource-manager/Microsoft.CognitiveServices/stable/2024-10-01/examples/PutRaiPolicy.json
if __name__ == "__main__":
    main()

To use the Azure SDK library in your project, see this documentation. To provide feedback on this code sample, open a GitHub issue

package armcognitiveservices_test

import (
	"context"
	"log"

	"github.com/Azure/azure-sdk-for-go/sdk/azcore/to"
	"github.com/Azure/azure-sdk-for-go/sdk/azidentity"
	"github.com/Azure/azure-sdk-for-go/sdk/resourcemanager/cognitiveservices/armcognitiveservices"
)

// Generated from example definition: https://github.com/Azure/azure-rest-api-specs/blob/069a65e8a6d1a6c0c58d9a9d97610b7103b6e8a5/specification/cognitiveservices/resource-manager/Microsoft.CognitiveServices/stable/2024-10-01/examples/PutRaiPolicy.json
func ExampleRaiPoliciesClient_CreateOrUpdate() {
	cred, err := azidentity.NewDefaultAzureCredential(nil)
	if err != nil {
		log.Fatalf("failed to obtain a credential: %v", err)
	}
	ctx := context.Background()
	clientFactory, err := armcognitiveservices.NewClientFactory("<subscription-id>", cred, nil)
	if err != nil {
		log.Fatalf("failed to create client: %v", err)
	}
	res, err := clientFactory.NewRaiPoliciesClient().CreateOrUpdate(ctx, "resourceGroupName", "accountName", "raiPolicyName", armcognitiveservices.RaiPolicy{
		Properties: &armcognitiveservices.RaiPolicyProperties{
			BasePolicyName: to.Ptr("Microsoft.Default"),
			ContentFilters: []*armcognitiveservices.RaiPolicyContentFilter{
				{
					Name:              to.Ptr("Hate"),
					Blocking:          to.Ptr(false),
					Enabled:           to.Ptr(false),
					SeverityThreshold: to.Ptr(armcognitiveservices.ContentLevelHigh),
					Source:            to.Ptr(armcognitiveservices.RaiPolicyContentSourcePrompt),
				},
				{
					Name:              to.Ptr("Hate"),
					Blocking:          to.Ptr(true),
					Enabled:           to.Ptr(true),
					SeverityThreshold: to.Ptr(armcognitiveservices.ContentLevelMedium),
					Source:            to.Ptr(armcognitiveservices.RaiPolicyContentSourceCompletion),
				},
				{
					Name:              to.Ptr("Sexual"),
					Blocking:          to.Ptr(true),
					Enabled:           to.Ptr(true),
					SeverityThreshold: to.Ptr(armcognitiveservices.ContentLevelHigh),
					Source:            to.Ptr(armcognitiveservices.RaiPolicyContentSourcePrompt),
				},
				{
					Name:              to.Ptr("Sexual"),
					Blocking:          to.Ptr(true),
					Enabled:           to.Ptr(true),
					SeverityThreshold: to.Ptr(armcognitiveservices.ContentLevelMedium),
					Source:            to.Ptr(armcognitiveservices.RaiPolicyContentSourceCompletion),
				},
				{
					Name:              to.Ptr("Selfharm"),
					Blocking:          to.Ptr(true),
					Enabled:           to.Ptr(true),
					SeverityThreshold: to.Ptr(armcognitiveservices.ContentLevelHigh),
					Source:            to.Ptr(armcognitiveservices.RaiPolicyContentSourcePrompt),
				},
				{
					Name:              to.Ptr("Selfharm"),
					Blocking:          to.Ptr(true),
					Enabled:           to.Ptr(true),
					SeverityThreshold: to.Ptr(armcognitiveservices.ContentLevelMedium),
					Source:            to.Ptr(armcognitiveservices.RaiPolicyContentSourceCompletion),
				},
				{
					Name:              to.Ptr("Violence"),
					Blocking:          to.Ptr(true),
					Enabled:           to.Ptr(true),
					SeverityThreshold: to.Ptr(armcognitiveservices.ContentLevelMedium),
					Source:            to.Ptr(armcognitiveservices.RaiPolicyContentSourcePrompt),
				},
				{
					Name:              to.Ptr("Violence"),
					Blocking:          to.Ptr(true),
					Enabled:           to.Ptr(true),
					SeverityThreshold: to.Ptr(armcognitiveservices.ContentLevelMedium),
					Source:            to.Ptr(armcognitiveservices.RaiPolicyContentSourceCompletion),
				},
				{
					Name:     to.Ptr("Jailbreak"),
					Blocking: to.Ptr(true),
					Enabled:  to.Ptr(true),
					Source:   to.Ptr(armcognitiveservices.RaiPolicyContentSourcePrompt),
				},
				{
					Name:     to.Ptr("Protected Material Text"),
					Blocking: to.Ptr(true),
					Enabled:  to.Ptr(true),
					Source:   to.Ptr(armcognitiveservices.RaiPolicyContentSourceCompletion),
				},
				{
					Name:     to.Ptr("Protected Material Code"),
					Blocking: to.Ptr(true),
					Enabled:  to.Ptr(true),
					Source:   to.Ptr(armcognitiveservices.RaiPolicyContentSourceCompletion),
				},
				{
					Name:     to.Ptr("Profanity"),
					Blocking: to.Ptr(true),
					Enabled:  to.Ptr(true),
					Source:   to.Ptr(armcognitiveservices.RaiPolicyContentSourcePrompt),
				}},
			Mode: to.Ptr(armcognitiveservices.RaiPolicyModeAsynchronousFilter),
		},
	}, nil)
	if err != nil {
		log.Fatalf("failed to finish the request: %v", err)
	}
	// You could use response here. We use blank identifier for just demo purposes.
	_ = res
	// If the HTTP response code is 200 as defined in example definition, your response structure would look as follows. Please pay attention that all the values in the output are fake values for just demo purposes.
	// res.RaiPolicy = armcognitiveservices.RaiPolicy{
	// 	Name: to.Ptr("raiPolicyName"),
	// 	ID: to.Ptr("/subscriptions/00000000-0000-0000-0000-000000000000/resourceGroups/resourceGroupName/providers/Microsoft.CognitiveServices/accounts/accountName/raiPolicies/raiPolicyName"),
	// 	Properties: &armcognitiveservices.RaiPolicyProperties{
	// 		BasePolicyName: to.Ptr("Microsoft.Default"),
	// 		ContentFilters: []*armcognitiveservices.RaiPolicyContentFilter{
	// 			{
	// 				Name: to.Ptr("Hate"),
	// 				Blocking: to.Ptr(false),
	// 				Enabled: to.Ptr(false),
	// 				SeverityThreshold: to.Ptr(armcognitiveservices.ContentLevelHigh),
	// 				Source: to.Ptr(armcognitiveservices.RaiPolicyContentSourcePrompt),
	// 			},
	// 			{
	// 				Name: to.Ptr("Hate"),
	// 				Blocking: to.Ptr(true),
	// 				Enabled: to.Ptr(true),
	// 				SeverityThreshold: to.Ptr(armcognitiveservices.ContentLevelMedium),
	// 				Source: to.Ptr(armcognitiveservices.RaiPolicyContentSourceCompletion),
	// 			},
	// 			{
	// 				Name: to.Ptr("Sexual"),
	// 				Blocking: to.Ptr(true),
	// 				Enabled: to.Ptr(true),
	// 				SeverityThreshold: to.Ptr(armcognitiveservices.ContentLevelHigh),
	// 				Source: to.Ptr(armcognitiveservices.RaiPolicyContentSourcePrompt),
	// 			},
	// 			{
	// 				Name: to.Ptr("Sexual"),
	// 				Blocking: to.Ptr(true),
	// 				Enabled: to.Ptr(true),
	// 				SeverityThreshold: to.Ptr(armcognitiveservices.ContentLevelMedium),
	// 				Source: to.Ptr(armcognitiveservices.RaiPolicyContentSourceCompletion),
	// 			},
	// 			{
	// 				Name: to.Ptr("Selfharm"),
	// 				Blocking: to.Ptr(true),
	// 				Enabled: to.Ptr(true),
	// 				SeverityThreshold: to.Ptr(armcognitiveservices.ContentLevelHigh),
	// 				Source: to.Ptr(armcognitiveservices.RaiPolicyContentSourcePrompt),
	// 			},
	// 			{
	// 				Name: to.Ptr("Selfharm"),
	// 				Blocking: to.Ptr(true),
	// 				Enabled: to.Ptr(true),
	// 				SeverityThreshold: to.Ptr(armcognitiveservices.ContentLevelMedium),
	// 				Source: to.Ptr(armcognitiveservices.RaiPolicyContentSourceCompletion),
	// 			},
	// 			{
	// 				Name: to.Ptr("Violence"),
	// 				Blocking: to.Ptr(true),
	// 				Enabled: to.Ptr(true),
	// 				SeverityThreshold: to.Ptr(armcognitiveservices.ContentLevelMedium),
	// 				Source: to.Ptr(armcognitiveservices.RaiPolicyContentSourcePrompt),
	// 			},
	// 			{
	// 				Name: to.Ptr("Violence"),
	// 				Blocking: to.Ptr(true),
	// 				Enabled: to.Ptr(true),
	// 				SeverityThreshold: to.Ptr(armcognitiveservices.ContentLevelMedium),
	// 				Source: to.Ptr(armcognitiveservices.RaiPolicyContentSourceCompletion),
	// 			},
	// 			{
	// 				Name: to.Ptr("Jailbreak"),
	// 				Blocking: to.Ptr(true),
	// 				Enabled: to.Ptr(true),
	// 				Source: to.Ptr(armcognitiveservices.RaiPolicyContentSourcePrompt),
	// 			},
	// 			{
	// 				Name: to.Ptr("Protected Material Text"),
	// 				Blocking: to.Ptr(true),
	// 				Enabled: to.Ptr(true),
	// 				Source: to.Ptr(armcognitiveservices.RaiPolicyContentSourceCompletion),
	// 			},
	// 			{
	// 				Name: to.Ptr("Protected Material Code"),
	// 				Blocking: to.Ptr(true),
	// 				Enabled: to.Ptr(true),
	// 				Source: to.Ptr(armcognitiveservices.RaiPolicyContentSourceCompletion),
	// 			},
	// 			{
	// 				Name: to.Ptr("Profanity"),
	// 				Blocking: to.Ptr(true),
	// 				Enabled: to.Ptr(true),
	// 				Source: to.Ptr(armcognitiveservices.RaiPolicyContentSourcePrompt),
	// 		}},
	// 		Mode: to.Ptr(armcognitiveservices.RaiPolicyModeAsynchronousFilter),
	// 	},
	// }
}

To use the Azure SDK library in your project, see this documentation. To provide feedback on this code sample, open a GitHub issue

const { CognitiveServicesManagementClient } = require("@azure/arm-cognitiveservices");
const { DefaultAzureCredential } = require("@azure/identity");

/**
 * This sample demonstrates how to Update the state of specified Content Filters associated with the Azure OpenAI account.
 *
 * @summary Update the state of specified Content Filters associated with the Azure OpenAI account.
 * x-ms-original-file: specification/cognitiveservices/resource-manager/Microsoft.CognitiveServices/stable/2024-10-01/examples/PutRaiPolicy.json
 */
async function putRaiPolicy() {
  const subscriptionId =
    process.env["COGNITIVESERVICES_SUBSCRIPTION_ID"] || "00000000-0000-0000-0000-000000000000";
  const resourceGroupName = process.env["COGNITIVESERVICES_RESOURCE_GROUP"] || "resourceGroupName";
  const accountName = "accountName";
  const raiPolicyName = "raiPolicyName";
  const raiPolicy = {
    properties: {
      basePolicyName: "Microsoft.Default",
      contentFilters: [
        {
          name: "Hate",
          blocking: false,
          enabled: false,
          severityThreshold: "High",
          source: "Prompt",
        },
        {
          name: "Hate",
          blocking: true,
          enabled: true,
          severityThreshold: "Medium",
          source: "Completion",
        },
        {
          name: "Sexual",
          blocking: true,
          enabled: true,
          severityThreshold: "High",
          source: "Prompt",
        },
        {
          name: "Sexual",
          blocking: true,
          enabled: true,
          severityThreshold: "Medium",
          source: "Completion",
        },
        {
          name: "Selfharm",
          blocking: true,
          enabled: true,
          severityThreshold: "High",
          source: "Prompt",
        },
        {
          name: "Selfharm",
          blocking: true,
          enabled: true,
          severityThreshold: "Medium",
          source: "Completion",
        },
        {
          name: "Violence",
          blocking: true,
          enabled: true,
          severityThreshold: "Medium",
          source: "Prompt",
        },
        {
          name: "Violence",
          blocking: true,
          enabled: true,
          severityThreshold: "Medium",
          source: "Completion",
        },
        { name: "Jailbreak", blocking: true, enabled: true, source: "Prompt" },
        {
          name: "Protected Material Text",
          blocking: true,
          enabled: true,
          source: "Completion",
        },
        {
          name: "Protected Material Code",
          blocking: true,
          enabled: true,
          source: "Completion",
        },
        { name: "Profanity", blocking: true, enabled: true, source: "Prompt" },
      ],
      mode: "Asynchronous_filter",
    },
  };
  const credential = new DefaultAzureCredential();
  const client = new CognitiveServicesManagementClient(credential, subscriptionId);
  const result = await client.raiPolicies.createOrUpdate(
    resourceGroupName,
    accountName,
    raiPolicyName,
    raiPolicy,
  );
  console.log(result);
}

To use the Azure SDK library in your project, see this documentation. To provide feedback on this code sample, open a GitHub issue

using Azure;
using Azure.ResourceManager;
using System;
using System.Threading.Tasks;
using Azure.Core;
using Azure.Identity;
using Azure.ResourceManager.CognitiveServices.Models;
using Azure.ResourceManager.CognitiveServices;

// Generated from example definition: specification/cognitiveservices/resource-manager/Microsoft.CognitiveServices/stable/2024-10-01/examples/PutRaiPolicy.json
// this example is just showing the usage of "RaiPolicies_CreateOrUpdate" operation, for the dependent resources, they will have to be created separately.

// get your azure access token, for more details of how Azure SDK get your access token, please refer to https://free.blessedness.top/en-us/dotnet/azure/sdk/authentication?tabs=command-line
TokenCredential cred = new DefaultAzureCredential();
// authenticate your client
ArmClient client = new ArmClient(cred);

// this example assumes you already have this CognitiveServicesAccountResource created on azure
// for more information of creating CognitiveServicesAccountResource, please refer to the document of CognitiveServicesAccountResource
string subscriptionId = "00000000-0000-0000-0000-000000000000";
string resourceGroupName = "resourceGroupName";
string accountName = "accountName";
ResourceIdentifier cognitiveServicesAccountResourceId = CognitiveServicesAccountResource.CreateResourceIdentifier(subscriptionId, resourceGroupName, accountName);
CognitiveServicesAccountResource cognitiveServicesAccount = client.GetCognitiveServicesAccountResource(cognitiveServicesAccountResourceId);

// get the collection of this RaiPolicyResource
RaiPolicyCollection collection = cognitiveServicesAccount.GetRaiPolicies();

// invoke the operation
string raiPolicyName = "raiPolicyName";
RaiPolicyData data = new RaiPolicyData()
{
    Properties = new RaiPolicyProperties()
    {
        Mode = RaiPolicyMode.AsynchronousFilter,
        BasePolicyName = "Microsoft.Default",
        ContentFilters =
        {
        new RaiPolicyContentFilter()
        {
        Name = "hate",
        Enabled = false,
        SeverityThreshold = RaiPolicyContentLevel.High,
        Blocking = false,
        Source = RaiPolicyContentSource.Prompt,
        },new RaiPolicyContentFilter()
        {
        Name = "hate",
        Enabled = true,
        SeverityThreshold = RaiPolicyContentLevel.Medium,
        Blocking = true,
        Source = RaiPolicyContentSource.Completion,
        },new RaiPolicyContentFilter()
        {
        Name = "sexual",
        Enabled = true,
        SeverityThreshold = RaiPolicyContentLevel.High,
        Blocking = true,
        Source = RaiPolicyContentSource.Prompt,
        },new RaiPolicyContentFilter()
        {
        Name = "sexual",
        Enabled = true,
        SeverityThreshold = RaiPolicyContentLevel.Medium,
        Blocking = true,
        Source = RaiPolicyContentSource.Completion,
        },new RaiPolicyContentFilter()
        {
        Name = "selfharm",
        Enabled = true,
        SeverityThreshold = RaiPolicyContentLevel.High,
        Blocking = true,
        Source = RaiPolicyContentSource.Prompt,
        },new RaiPolicyContentFilter()
        {
        Name = "selfharm",
        Enabled = true,
        SeverityThreshold = RaiPolicyContentLevel.Medium,
        Blocking = true,
        Source = RaiPolicyContentSource.Completion,
        },new RaiPolicyContentFilter()
        {
        Name = "violence",
        Enabled = true,
        SeverityThreshold = RaiPolicyContentLevel.Medium,
        Blocking = true,
        Source = RaiPolicyContentSource.Prompt,
        },new RaiPolicyContentFilter()
        {
        Name = "violence",
        Enabled = true,
        SeverityThreshold = RaiPolicyContentLevel.Medium,
        Blocking = true,
        Source = RaiPolicyContentSource.Completion,
        },new RaiPolicyContentFilter()
        {
        Name = "jailbreak",
        Enabled = true,
        Blocking = true,
        Source = RaiPolicyContentSource.Prompt,
        },new RaiPolicyContentFilter()
        {
        Name = "protected_material_text",
        Enabled = true,
        Blocking = true,
        Source = RaiPolicyContentSource.Completion,
        },new RaiPolicyContentFilter()
        {
        Name = "protected_material_code",
        Enabled = true,
        Blocking = true,
        Source = RaiPolicyContentSource.Completion,
        },new RaiPolicyContentFilter()
        {
        Name = "profanity",
        Enabled = true,
        Blocking = true,
        Source = RaiPolicyContentSource.Prompt,
        }
        },
    },
};
ArmOperation<RaiPolicyResource> lro = await collection.CreateOrUpdateAsync(WaitUntil.Completed, raiPolicyName, data);
RaiPolicyResource result = lro.Value;

// the variable result is a resource, you could call other operations on this instance as well
// but just for demo, we get its data from this resource instance
RaiPolicyData resourceData = result.Data;
// for demo we just print out the id
Console.WriteLine($"Succeeded on id: {resourceData.Id}");

To use the Azure SDK library in your project, see this documentation. To provide feedback on this code sample, open a GitHub issue

Sample response

Status code:: 200

{
  "name": "raiPolicyName",
  "id": "/subscriptions/00000000-0000-0000-0000-000000000000/resourceGroups/resourceGroupName/providers/Microsoft.CognitiveServices/accounts/accountName/raiPolicies/raiPolicyName",
  "properties": {
    "basePolicyName": "Microsoft.Default",
    "mode": "Asynchronous_filter",
    "contentFilters": [
      {
        "name": "Hate",
        "blocking": false,
        "enabled": false,
        "severityThreshold": "High",
        "source": "Prompt"
      },
      {
        "name": "Hate",
        "blocking": true,
        "enabled": true,
        "severityThreshold": "Medium",
        "source": "Completion"
      },
      {
        "name": "Sexual",
        "blocking": true,
        "enabled": true,
        "severityThreshold": "High",
        "source": "Prompt"
      },
      {
        "name": "Sexual",
        "blocking": true,
        "enabled": true,
        "severityThreshold": "Medium",
        "source": "Completion"
      },
      {
        "name": "Selfharm",
        "blocking": true,
        "enabled": true,
        "severityThreshold": "High",
        "source": "Prompt"
      },
      {
        "name": "Selfharm",
        "blocking": true,
        "enabled": true,
        "severityThreshold": "Medium",
        "source": "Completion"
      },
      {
        "name": "Violence",
        "blocking": true,
        "enabled": true,
        "severityThreshold": "Medium",
        "source": "Prompt"
      },
      {
        "name": "Violence",
        "blocking": true,
        "enabled": true,
        "severityThreshold": "Medium",
        "source": "Completion"
      },
      {
        "name": "Jailbreak",
        "blocking": true,
        "source": "Prompt",
        "enabled": true
      },
      {
        "name": "Protected Material Text",
        "blocking": true,
        "source": "Completion",
        "enabled": true
      },
      {
        "name": "Protected Material Code",
        "blocking": true,
        "source": "Completion",
        "enabled": true
      },
      {
        "name": "Profanity",
        "blocking": true,
        "source": "Prompt",
        "enabled": true
      }
    ]
  }
}

Status code:: 201

{
  "name": "raiPolicyName",
  "id": "/subscriptions/00000000-0000-0000-0000-000000000000/resourceGroups/resourceGroupName/providers/Microsoft.CognitiveServices/accounts/accountName/raiPolicies/raiPolicyName",
  "properties": {
    "basePolicyName": "Microsoft.Default",
    "mode": "Asynchronous_filter",
    "contentFilters": [
      {
        "name": "Hate",
        "blocking": false,
        "enabled": false,
        "severityThreshold": "High",
        "source": "Prompt"
      },
      {
        "name": "Hate",
        "blocking": true,
        "enabled": true,
        "severityThreshold": "Medium",
        "source": "Completion"
      },
      {
        "name": "Sexual",
        "blocking": true,
        "enabled": true,
        "severityThreshold": "High",
        "source": "Prompt"
      },
      {
        "name": "Sexual",
        "blocking": true,
        "enabled": true,
        "severityThreshold": "Medium",
        "source": "Completion"
      },
      {
        "name": "Selfharm",
        "blocking": true,
        "enabled": true,
        "severityThreshold": "High",
        "source": "Prompt"
      },
      {
        "name": "Selfharm",
        "blocking": true,
        "enabled": true,
        "severityThreshold": "Medium",
        "source": "Completion"
      },
      {
        "name": "Violence",
        "blocking": true,
        "enabled": true,
        "severityThreshold": "Medium",
        "source": "Prompt"
      },
      {
        "name": "Violence",
        "blocking": true,
        "enabled": true,
        "severityThreshold": "Medium",
        "source": "Completion"
      },
      {
        "name": "Jailbreak",
        "blocking": true,
        "source": "Prompt",
        "enabled": true
      },
      {
        "name": "Protected Material Text",
        "blocking": true,
        "source": "Completion",
        "enabled": true
      },
      {
        "name": "Protected Material Code",
        "blocking": true,
        "source": "Completion",
        "enabled": true
      },
      {
        "name": "Profanity",
        "blocking": true,
        "source": "Prompt",
        "enabled": true
      }
    ]
  }
}

Definitions

Name	Description
ContentLevel	Level at which content is filtered.
createdByType	The type of identity that created the resource.
CustomBlocklistConfig	Gets or sets the source to which filter applies.
ErrorAdditionalInfo	The resource management error additional info.
ErrorDetail	The error detail.
ErrorResponse	Error response
RaiPolicy	Cognitive Services RaiPolicy.
RaiPolicyContentFilter	Azure OpenAI Content Filter.
RaiPolicyContentSource	Content source to apply the Content Filters.
RaiPolicyMode	Rai policy mode. The enum value mapping is as below: Default = 0, Deferred=1, Blocking=2, Asynchronous_filter =3. Please use 'Asynchronous_filter' after 2024-10-01. It is the same as 'Deferred' in previous version.
RaiPolicyProperties	Azure OpenAI Content Filters properties.
RaiPolicyType	Content Filters policy type.
systemData	Metadata pertaining to creation and last modification of the resource.

ContentLevel

Enumeration

Level at which content is filtered.

Value	Description
Low
Medium
High

createdByType

Enumeration

The type of identity that created the resource.

Value	Description
User
Application
ManagedIdentity
Key

CustomBlocklistConfig

Object

Gets or sets the source to which filter applies.

Name	Type	Description
blocking	boolean	If blocking would occur.
blocklistName	string	Name of ContentFilter.
source	RaiPolicyContentSource	Content source to apply the Content Filters.

ErrorAdditionalInfo

Object

The resource management error additional info.

Name	Type	Description
info	object	The additional info.
type	string	The additional info type.

ErrorDetail

Object

The error detail.

Name	Type	Description
additionalInfo	ErrorAdditionalInfo[]	The error additional info.
code	string	The error code.
details	ErrorDetail[]	The error details.
message	string	The error message.
target	string	The error target.

ErrorResponse

Object

Error response

Name	Type	Description
error	ErrorDetail	The error object.

RaiPolicy

Object

Cognitive Services RaiPolicy.

Name	Type	Description
etag	string	Resource Etag.
id	string	Fully qualified resource ID for the resource. Ex - /subscriptions/{subscriptionId}/resourceGroups/{resourceGroupName}/providers/{resourceProviderNamespace}/{resourceType}/{resourceName}
name	string	The name of the resource
properties	RaiPolicyProperties	Properties of Cognitive Services RaiPolicy.
systemData	systemData	Metadata pertaining to creation and last modification of the resource.
tags	object	Resource tags.
type	string	The type of the resource. E.g. "Microsoft.Compute/virtualMachines" or "Microsoft.Storage/storageAccounts"

RaiPolicyContentFilter

Object

Azure OpenAI Content Filter.

Name	Type	Description
blocking	boolean	If blocking would occur.
enabled	boolean	If the ContentFilter is enabled.
name	string	Name of ContentFilter.
severityThreshold	ContentLevel	Level at which content is filtered.
source	RaiPolicyContentSource	Content source to apply the Content Filters.

RaiPolicyContentSource

Enumeration

Content source to apply the Content Filters.

Value	Description
Prompt
Completion

RaiPolicyMode

Enumeration

Rai policy mode. The enum value mapping is as below: Default = 0, Deferred=1, Blocking=2, Asynchronous_filter =3. Please use 'Asynchronous_filter' after 2024-10-01. It is the same as 'Deferred' in previous version.

Value	Description
Default
Deferred
Blocking
Asynchronous_filter

RaiPolicyProperties

Object

Azure OpenAI Content Filters properties.

Name	Type	Description
basePolicyName	string	Name of Rai policy.
contentFilters	RaiPolicyContentFilter[]	The list of Content Filters.
customBlocklists	CustomBlocklistConfig[]	The list of custom Blocklist.
mode	RaiPolicyMode	Rai policy mode. The enum value mapping is as below: Default = 0, Deferred=1, Blocking=2, Asynchronous_filter =3. Please use 'Asynchronous_filter' after 2024-10-01. It is the same as 'Deferred' in previous version.
type	RaiPolicyType	Content Filters policy type.

RaiPolicyType

Enumeration

Content Filters policy type.

Value	Description
UserManaged
SystemManaged

systemData

Object

Metadata pertaining to creation and last modification of the resource.

Name	Type	Description
createdAt	string (date-time)	The timestamp of resource creation (UTC).
createdBy	string	The identity that created the resource.
createdByType	createdByType	The type of identity that created the resource.
lastModifiedAt	string (date-time)	The timestamp of resource last modification (UTC)
lastModifiedBy	string	The identity that last modified the resource.
lastModifiedByType	createdByType	The type of identity that last modified the resource.