John Hearty

John Hearty

Vice President, AI Governance Mastercard
John Hearty

Main Conference Day Two - November 5, 2026

8:35 AM Case Study: Human-in-the-Loop Architectures: Designing Control, Trust, and Intervention in Agentic Systems

As organisations introduce more autonomous systems into critical workflows, fully removing humans from the loop is rarely realistic, or desirable. Instead, leading organisations are designing hybrid models where agents operate independently within defined boundaries, with human oversight, escalation, and intervention built in. The challenge is determining where and how humans should remain involved to ensure trust, control, and clear accountability, without undermining the efficiency gains agentic AI promises.


• Define where human oversight is required across decision-making, execution, and escalation points
• Establish clear accountability models—who is responsible for agent decisions, outcomes, and failures
• Design intervention mechanisms that allow humans to step in without disrupting system performance
• Build trust through transparency, explainability, and controlled autonomy in high-stakes environments

Check out the incredible speaker line-up to see who will be joining John.

Download The Latest Agenda