When an AI Agent Acts, Who Is Accountable?
- Date: Wednesday, 25 November 2026
- Time: 3:00 – 7:30 pm
- Venue: KPMG Frankfurt, The Squaire, Am Flughafen 12, 60549 Frankfurt am Main
- Language: German
- Admission: Free of charge, limited number of places
AI agents access data and systems on their own, use tools and APIs, and carry out processes. The risk shifts from a wrong answer to an uncontrolled action. Organizations need to define what an agent may do, under which conditions, and how transparency and traceability are ensured, which is crucial for information security, compliance, and data protection.
At the same time, AI is one of the biggest levers for Identity & Access Management itself: at Bayer, the AI IAM Workbench handles several thousand access requests every day, and AI agents implement 98.6% of them automatically.
Together with SailPoint, KPMG invites you to an exclusive Executive Briefing that brings both perspectives together: using AI effectively in IAM, and governing AI identities securely.
What to expect
- A view from the pharma industry: Bayer and KPMG share how AI agents turn ServiceNow access requests into SailPoint actions, and what has proven viable in a regulated, international environment.
- Governance and control for AI identities: Transparency, clear accountability, least privilege, binding policies, human-in-the-loop, and lifecycle management.
- Beyond classic IAM models: Why role-based concepts reach their limits with autonomous identities, and how policy-based access control, delegated permissions, and just-in-time authorization can close the gap.
- Roundtable: Discuss the AI IAM operating model with the speakers and Prof. Dr. Dirk Loomans (Mainz University of Applied Sciences), and find out which areas to tackle first.
Speakers: Dr. Florian Kohlar and Ralf Eduard Defort (KPMG), Felix Schröder (Bayer), Prof. Dr. Dirk Loomans (Mainz University of Applied Sciences), Matt Brunn and Christian Bersch (SailPoint)
Please note: this event will be held in German.

