Microsoft releases two open-source tools intended to improve security practices in agentic AI development. The company introduces RAMPART, a continuous testing and risk-assessment framework for AI agents, and Clarity, a structured design review tool. Microsoft says the tools come from its AI Red Team, an internal group that stress-tests Microsoft’s own AI systems, and that both tools were used internally before being published. RAMPART is built on top of PyRIT, Microsoft’s existing open-source red-teaming library, and is designed to integrate into developer workflows by fitting into a Pytest-native testing approach. The framework provides mechanisms for writing and running safety and security tests for AI agents. Clarity focuses on bringing more discipline to agent design by supporting structured review processes that aim to identify and address potential safety and security issues earlier in development. Across the reports, the common theme is that Microsoft is offering tooling to help developers evaluate agent behavior and reduce risk during building and iteration, rather than treating security as something addressed only after deployment.
Microsoft open-sources RAMPART and Clarity for security testing of AI agents
Microsoft releases two open-source tools intended to improve security practices in agentic AI development. The company introduces RAMPART, a continuous testing and risk-assessment framework for AI age...
- Microsoft open-sources two tools for agentic AI security: RAMPART and Clarity.
- RAMPART is a testing framework built on PyRIT and designed to work with Pytest-native workflows.
- Clarity provides a structured design review process for safer agent development.
- Both tools originate from Microsoft’s AI Red Team and were used internally before being released.
- The releases aim to support ongoing safety and security assessment during development of AI agents.
Microsoft has open-sourced two tools aimed at bringing security discipline to AI agent development: Clarity, a structured design review tool, and RAMPART, a continuous testing framework. The release comes from Microsoft’s AI Red Team, the company’s internal unit that stress-tests its own AI systems, and both tools have been used internally before being open-sourced. RAMPART: A test harness RAMPART is built on top of PyRIT, Microsoft’s existing open-source red-teaming library, and is designed to slot … More → The post Microsoft open-sources tools for designing and testing AI agents appeared first on Help Net Security.
3 months agoRedmond open sources two tools for building and maintaining safer agents
3 months agoMicrosoft has unveiled two new open-source tools called RAMPART and Clarity to assist developers in better testing the security of artificial intelligence (AI) agents. RAMPART, short for Risk Assessment and Measurement Platform for Agentic Red Teaming, functions as a Pytest-native safety and security testing framework for writing and running safety and security tests for AI agents, covering
3 months agoMeta agrees to $17–$18 billion teen-safety settlement, sets default time limits on Instagram and Facebook
Meta has reached a proposed settlement with a coalition of U.S. state attorneys general over allegations that Instagram...
Apple launches new Mac mini and Mac Studio models with M6, M5 Max and M5 Ultra
Apple is taking pre-orders for updated Mac mini and Mac Studio desktops, with both product lines shipping to buyers on S...
Nvidia agrees to acquire Hugging Face in reported $12.9 billion AI deal
Nvidia is reported to have agreed to acquire Hugging Face for about $12.9 billion, according to multiple outlets. The de...