Welcome to today's video presentation, where we will be exploring AI visibility within Cisco AI defense. This solution integrates with Amazon Web Services, Azure AI, and Google Cloud Platform.00:04Providing enhanced security and operational insights, let's delve into the features and benefits of this integration. We'll begin by outlining today's agenda. 1st, we will provide an overview of AI visibility and discuss how the AI inventory feature enhances AWS integration. Next, we'll demonstrate how to configure AI.00:16Inventory and examine the integration of AWS. We will explore the cloud formation stack and learn how to view AI inventory.00:36Finally, we'll discuss best practices for maximizing visibility and ensuring optimal security. To start, let's define AI visibility. This functionality offers an automated, consolidated discovery of AI assets across AWS, Azure, and GCP.00:43As we focus on AWS, visibility includes discovery and inventory changes to AI assets such as bedrock models and sagemaker models directly into AI defects. It identifies both sanctioned and unsanctioned AI workloads on public and private clouds. AI visibility closes visibility gaps in AI infrastructure by continuously discovering.01:01AI assets that traditional security measures might miss. This comprehensive inventory enables proactive security management.01:21Validation of AI assets and provides a clear understanding of asset usage and activity context. Once logged into the Cisco Security Cloud Control, navigate to the AI defense section from the left hand menu. This section provides tools for managing and monitoring AI assets within the system.01:28We will use this area to begin integrating our AWS resources. Within AI defense, access the administration menu and select integrations. Here, you'll find two options. The cloud tab for integrating cloud services like AWS and the others tab for additional solutions such as secure access and Splunk. Today.01:46Hey, we'll focus on the cloud tab for AWS integration.02:06Now, under the cloud tab, initiate the integration wizard by selecting the add integration button and choose Amazon Web Services. This launch sequence will guide you through creating a cloud formation stack in AWS necessary for connecting and securing your AI assets efficiently.02:11In step one of the integration process, configure the connection by providing AWS details. Start with a display name relevant to your AWS organization or accounts. Choose between organizational unit for entire organizations or account for individual accounts. In this example, we'll select account and enter the account ID.02:29Then we'll select the capability for AI defense to perform actions on discovered assets by selecting the bottom feature. The Read Cloud trail events option may also be done for near real time inventory updates of AI assets by reviewing cloud trail logs, However, that will be covered on another video.02:47Proceed to step two to deploy IAM roles. Initiate a cloud formation instance to create the necessary IAM roles and features. Click on the launch cloud formation in AWS button to begin defining the stack, which is crucial for creating the two IAM roles and specific actions used by AI defense. The bottom two fields will be filled later from the cloud form.03:05Information stack output with specific arns from the two IAM roles created.03:25We'll come back to this later. During the quick create stack setup, two significant roles are established. The discovery role and the action role. The discovery can perform AI model discovery and extended AI discovery. The action role can perform AI model scanning.03:29AI validation, AI runtime monitoring, and read cloud trail events. The existing options details are optional for customization, but you may stick to the default settings to simplify the quick create stack setup. Scroll down to the acknowledgement section to complete the process and create the stack.03:46As we reach the acknowledgment section, this action will create IAM resources. Check the acknowledgement box and then click the create stack button to initiate the creation of the cloud formation stack. This process is generally swift taking NO longer than 2 min. Upon successfully creating the cloud formation stack, you should see the status.04:04Marked as create complete on the left side of the screen.04:24This indicates that the stack was generated without issues. At this point, you will navigate to the outputs tab among the visible options. This tab is essential as it displays the values you'll need to copy and paste back into AI defense. Carefully note these values for the subsequent steps. Now, let's focus on the specific outputs from the stack.04:27You'll need to copy two key values from this output.04:47The action role ARN and the discovery role ARN, the action role ARN, highlighted in green, must be transferred to the AI defense Asset action role ARN field. Similarly, the discovery role ARN highlighted in blue, should be inputted into the AI defense asset discovery role, ARN field.04:49After entering these values, click on Complete to conclude the AWS integration process. Post integration, you will be redirected back to the integrations page. The API synchronization takes approximately 1 min during this time.05:08Refresh the page until the status of your AWS integration updates to active. This confirms a successful connection and enables AI defense to effectively utilize AWS services. To take advantage of the populated data from AWS, navigate to AI assets in the left hand menu.05:23And then proceed to AI inventory. This interface displays a comprehensive list of AI assets, including agents, models, endpoints, agent runtimes, and knowledge bases. Please note that the AI inventory updates daily, but for near real time insights, revisit the AWS integration and enable the cloud trail events capability.05:41Which facilitates detailed tracking and management of AI assets. Here are some recommended best practices for maximizing the AI inventory feature for your AWS integration with Cisco AI defense. One, leverage automated discovery.06:01This ensures complete AI asset visibility across all AWS accounts within the organization. Two, regularly review AI inventory. This account for an accurate record of AI assets activity tracking. Three, review IAM roles and permissions. Follow the principle of least privilege to secure AWS resource access. Four.06:17Apply consistent tagging. Use tags for efficient asset management and policy enforcement. Five, monitor inventory changes and anomalies, proactively identify security risks from unauthorized patterns and deployments. Six, enable AWS cloud trail integration.06:37This captures near real time inventory updates of AWS AI assets. If deploying the cloud trail integration, you can reuse existing deployed resources to reduce AWS costs. This brings us to the conclusion of our demonstration on using AI inventory with AWS integration as you continue exploring.06:52Further integrations.07:12You will unlock additional features and capabilities to safeguard AI assets against emerging threats. It's easier to protect what you have visibility into. We thank you for entrusting Cisco with your security needs and supporting our vision for advanced agenc security.07:13
