Finalist
GenAI Berlin Award 2026
TrustX Agent Risk Classification Framework
A risk classification framework for enterprises to evaluate their agents against and produce measurable controls against international AI governance standards.
Public Sector, Education & Social Good
About the project
Organizations are putting AI agents into production that can plan, use tools, and act with little human oversight. However, existing AI governance frameworks, such as the EU AI Act and NIST AI RMF, are not designed to risk-assess agentic AI systems.
The Responsible AI Institute (RAI) built the TrustX Agent Risk Classification Framework (ARC), which helps enterprises classify internal AI risk and translate governance requirements into clear, measurable controls. It is based on internationally recognised standards, including the EU AI Act, NIST AI Risk Management Framework and ISO/IEC 42001, and is designed to support organisations deploying AI at scale.
Since its publication, ARC has been widely used as an openly published resource and has been embedded into internal risk assessment systems of organizations across various industries, such as financial services and healthcare. Its creation launched two sector working groups within RAI: TrustX for Finance and TrustX for Healthcare.
Key details
Track
Public Sector, Education & Social Good
Organisation
Status
Finalist
In the team's words
Entering the GenAI Award lets us share a resource we built to address the gap in agentic AI evaluation with enterprise leaders across industries. As a non-profit, the Responsible AI Institute (RAI) built ARC as an open-source resource to help govern agentic AI before it does more harm than good. We noticed that existing AI governance frameworks were not designed for agentic systems capable of planning and acting on their own, and yet agents are being rapidly integrated into high-touch settings. As a result, we built ARC so that any organization can run a rigorous, standards-grounded risk assessment of an AI agent before it is deployed. In an age where AI is growing faster than any individual human can keep up, leveraging collective intelligence and making resources openly available is arguably one of the only ways we can work toward a more equitable and safe future in AI.
Hannah Liu
Research Engineer @ Responsible AI Institute
Europe’s go-to conference for GenAI leaders and enthusiasts
Attend GenAI Berlin to stay at the forefront of Generative AI, connect with the minds shaping the technology’s future, and explore its real-world impact across industries.
