AI Security, a new add-on for Azure
Firewall Premium, is now in public preview. Announced at Microsoft Ignite, it
introduces five capabilities to help you protect AI applications, safeguard
sensitive data, strengthen malware and application defenses, and simplify
firewall policy management.
Prompt injection protection
uses Azure AI Content Safety Prompt Shields to detect and block
prompt-injection attempts in selected traffic, without per-application SDK
integration.
Network DLP
integrates with Microsoft Purview to help prevent sensitive corporate data from
being uploaded to destinations your organization does not sanction, including
AI applications and models.
Malware detection
and application vulnerability protection bring purpose-built small
language models (SLMs), fine-tuned for network security and optimized to run on
CPUs, into inline inspection. These models analyze binary files and application
requests in transit to detect malware and exploit attempts, enabling policy
enforcement before identified threats reach their destinations.
AI policy
combines firewall configuration with observed network traffic to provide
contextual insights and recommendations. Use these insights to review unused
rules, identify opportunities to improve rule order, and investigate traffic
anomalies. Review recommendations and validate changes with a dry-run before
applying them.
Together, these capabilities bring
AI-powered intelligence to network protection and policy management.
Learn more: