Agentic AI in Financial Risk Management: Use Cases, Regulatory Exposure and the Question of Authority

Zurich Chapter: AI agents in financial services are now capable of directly supporting business processes, including acting on client accounts, and managing portfolios and risk systems. Agentic risk stems largely from the architecture of agent models which lack the same controls built for static models.  

This session focuses on the management of Agentic AI models to ensure AI governance is built to identify which agents are acting, define parameters for the work the models can do, and build accountability for the actions taken by AI Agents.  

Join the GARP Zurich Chapter and Professor Semyon Malamud and Elena Maran, for an insightful discussion of Agentic AI use cases and their implications for risk management and AI regulation.  

Join us to network with industry leaders and learn: 

• Where agentic AI is already reshaping risk management in financial institutions. 
• Why privilege escalation, behavioural drift and accountability gaps escape traditional controls. 
• How the EU AI Act, DORA, and adjacent EU rules apply when requirements are triggered by actions taken by AI agents at runtime. 
• How to move from permission to accountable authority, with an auditable record of every decision. 

Agenda

5:30 – 6:00 pm: Registration 

6:00 – 6:05 pm: Welcome remarks 

6:05 – 7:25 pm: Presentation/Panel discussion with Audience Q&A 

7:25 – 8:30 pm: Networking reception