As money and assets become programmable and financial activity moves toward 24/7 execution, traditional control models built around reviewing, reconciling and monitoring transactions after the fact will increasingly be insufficient. Identity, permissions, compliance, risk parameters and transaction policies will need to operate in real time—and, in some cases, be embedded directly into the transaction itself. With AI agents capable of initiating and executing financial activity added into the mix, the governance challenge becomes even more significant. This working group will examine how institutions redesign trust and control for programmable finance, determining who and what is authorized to transact, under which rules, with what degree of autonomy—and how accountability is maintained when transactions execute automatically.
Discussion points the working group will take up include: