Focus
Focus
Table of Contents

AI Runtime Security

AI Runtime Security leverages critical anomaly detection capabilities and protects AI models from manipulation to ensure the reliability and integrity of AI output data.
Palo Alto Networks AI Runtime Security is a purpose-built firewall to discover, protect, and defend the enterprise traffic flows against all potential threats focusing on addressing AI-specific vulnerabilities such as prompt injection, and denial-of-service attacks on AI models. It combines continuous runtime threat analysis of your AI applications, models, and data sets with AI powered security to stop attackers in their tracks. The AI Runtime Security leverages real-time AI-powered security protecting your AI application ecosystem from both AI-specific and conventional network attacks.
AI Runtime Security leverages critical anomaly detection capabilities and protects AI models from manipulation to ensure the reliability and integrity of AI output data. It rejects prompt injections, malicious responses, training data poisoning, malicious URLs, command and control, embedded unsafe URLs, and lateral threat movement.
AI Runtime Security uses Palo Alto Networks Strata Cloud Manager (SCM) as the main configuration and management engine. To begin with, activate and onboard your cloud service provider account on SCM. The AI Security Profile imports security capabilities from Enterprise DLP and URL Filtering for inline detection of threats in AI application traffic.
The AI Runtime Security is powered by the following four key elements:
Discover - The AI Runtime Security discovers your enterprise AI application and all other applications. The AI Runtime Security dashboard provides complete visibility and security insights of your AI and other applications in just a few clicks. You can effortlessly gain actionable intelligence on AI traffic flows covering your applications, models, user access, and infrastructure threats.
Deploy - The AI Runtime Security deployment using Terraform templates automates the deployment procedure reducing the human error, lowering the required time for manual configuration tasks, and for protecting your enterprise AI applications. Deploy your AI Runtime Security instance downloading the Terraform templates and provide permissions to your cloud service provider account projects to analyze flow logs and DNS logs.
Detect - Identify unprotected traffic flows with potential security threats to the cloud network and detect the potential security risks based on logs and recommended actions to remediate.
Defend - Shield your organization’s AI application ecosystem from AI-specific and conventional network attacks by leveraging real-time AI-powered security. Get the continuous discovery of the AI network traffic on the containers and namespaces.
To learn more about AI Runtime Security activation, onboarding, and deployment, see AI Runtime Security documentation.