Why don't you like this?
4:00 PM CEST, 10:00 AM EDT, 7:00 AM PDT
Who Gave the Agent Permission? Delegated Authority, Trust & Accountability in Agentic AI
- understand how delegated authority changes when AI agents act on behalf of humans.
- distinguish between agent ownership, the principal granting authority, and the agent’s own permissions.
- explore how to scope and constrain an agent’s effective authority.
- learn why traditional service-account and human-centric identity models are insufficient for autonomous agents.
- discover how signed, scoped, and revocable mandates can strengthen agent governance.
- see how authorization evidence can connect individual agent actions back to the authority behind them.
- prepare for audit, compliance, and accountability requirements in an increasingly agentic enterprise.
AI agents are moving from assisting users to independently accessing systems, calling tools, and taking consequential actions on our behalf. Traditional identity models were not designed for this shift. Organizations need new approaches to delegated authority that establish whose permissions an agent is exercising, constrain what it can do, and provide verifiable evidence of who authorized its actions.
Jonathan Care Practice Lead AI at KuppingerCole will examine how identity and access management must evolve as AI agents become autonomous actors. They will explore the distinction between agent ownership and delegated authority, why conventional human and service-account models fall short, and how organizations can establish clear boundaries, governance, and accountability for agents acting on behalf of others.
Grace Rachmany, ExecutiveDirector at Decentralized Identity Foundation (DIF) will explore how delegated authority can be implemented and demonstrated in practice. They will address scoped and revocable mandates, binding agents to verified principals, and creating an evidence chain that connects an agent’s actions to its authorization. This allows organizations to establish not only what happened, but who authorized it and within what limits.
Dr. Angelika Steinacker is a seasoned Cybersecurity professional with over 30 years of experience, including more than two decades specializing in Identity and Access Management (IAM). She has held leadership, executive consulting, and strategic advisory roles, most recently serving as CTO for IAM at IBM’s Consulting EMEA unit, where she focused on GenAI and IAM integration, Security for AI, Identity Fabric, and Zero Trust. She continues to advance IAM and Security focusing on strategic initiatives and innovation in these evolving domains.
Who Should Attend
This webinar is designed for CISOs, IAM and identity architects, security leaders, GRC professionals, and IT decision-makers responsible for governing AI agents, non-human identities, access, and emerging agentic systems.
