REGISTER HERE

• Date & Time: 20 October 2025, 1:30 - 5:00 PM
• Location: TBD

Intended Audience:

AI and Security Architecture and security leaders (managers and ICs with a security business focus)

Goal:

Develop a common reusable architecture for Agentic AI systems that is FINOS AI Governance Framework aligned. 

Why get involved:

As financial services adopt increasingly agentic AI systems, building secure, trustworthy, and scalable reference architectures is critical to establish common and practical patterns. The Fintech Open Source Foundation (FINOS) is uniquely positioned to convene industry experts and practitioners to collaboratively design these architectures.

In the context of the AI Readiness Special Interest Group (SIG), this Tabletop Exercise (TTX), led by our trusted security partner and FINOS member ControlPlane, will pivot from simulating cyberattacks to a fast-paced, technical working session focused on co-developing a hardened, FSI-tailored blueprint for agentic AI system design.

We will bridge the gap between functions and FINOS projects (Governance ->  AI Governance Framework; Engineering -> Common Cloud Controls; Architecures ->  Common Architecture Language Model), deriving an end-to-end reference architecture for agentic AI systems. This will span data, models, inference, agents/MCP, and serving, with identified controls, enforcement points, and guardrails at each stage.

If successful, the outcome will be a tangible, community-built, secure-by-design blueprint that institutions can adopt as part of their broader system and enterprise architectures. The AI Readiness SIG will maintain and evolve this blueprint, with ongoing contributions from FINOS members.

Useful links:

https://www.finos.org/ai-readiness

https://air-governance-framework.finos.org/

Limited space available