BMK Lakshminarayanan

You can’t govern autonomous systems simply by piling on approval gates. What happens when an AI agent makes decisions and takes action faster than your organisation can review them—and who owns the consequences when it gets something wrong? This session explores AI agents through a governance, risk and compliance (GRC) lens. We’ll examine how to design accountability into agentic systems from the outset: clear ownership, defined operating boundaries, traceable decisions, and tamper-evident audit trails. The goal is not merely to record what happened, but to establish who authorised the action, whether it stayed within its mandate, and who is responsible for responding when it did not. The central question: how do we enable autonomous action without creating an accountability gap?