Microsoft Releases New AI Security Tools

Microsoft has open-sourced two new tools, Rampart and Clarity, to help developers and security teams test and secure AI agents.

Microsoft has released two new open-source tools, Rampart and Clarity, designed to help developers and security teams improve the safety of AI agents. The tools aim to integrate security testing from the earliest stages of development, supporting a shift towards continuous safety engineering. According to Ram Shankar Siva Kumar, founder of Microsoft's AI Red Team, AI safety must evolve from a periodic check into an ongoing discipline. By providing practical, open tools, the company hopes to empower developers to build safety directly into their AI systems from the very beginning.

The release addresses the growing security challenges posed by the evolution of AI agents. As these systems move beyond simple chatbot functions to possess real operational authority, they become more vulnerable to new attacks like prompt injection and unintended autonomous actions. Rampart and Clarity provide a structured framework for identifying and mitigating these risks before deployment. This proactive approach is crucial for any organization building with agentic AI, helping ensure that autonomous systems behave as intended and do not introduce unforeseen vulnerabilities into their operational environments. This move reflects a wider industry trend toward prioritizing AI safety as capabilities expand.

Microsoft is providing free, open-source tools to tackle emerging security risks in agentic AI, like prompt injection, helping standardize safety practices in a rapidly growing field.

Businesses developing or deploying AI agents can use these tools to reduce security risks, build more reliable products, and increase customer trust by proactively addressing vulnerabilities before they are exploited.