Cisco has announced AI Defense, a comprehensive solution to secure enterprise use and development of AI applications, launching in March. It provides visibility into AI app usage, access controls for unsanctioned tools, and protection against threats like data leakage and malicious downloads. For building AI apps, it offers discovery of shadow and sanctioned apps, automated testing to identify vulnerabilities, and runtime protection against threats like prompt injection, DoS attacks, and sensitive data leaks. This end-to-end solution addresses risks from data exposure and vulnerabilities in AI systems, combining AI Access, Cloud Visibility, Model Validation, and Runtime Protection components.

‍