Bank Policy Institute: AI agents raise new questions around liability and consumer protection

Rep. Bill Foster (D-IL): “How do you think we should be thinking about the transition to agentic finances, where most transactions happen not between consumers and businesses or business to businesses, but through agents, and the choice of the exact payment rails may be sort of a side issue to these two agents that have already negotiated identity and authorization, and all these other things that you struggle with all the time, that will sort of be built in automatically to the agent to agent interaction? I worry about that, whether we were ready with the legal certainty of what happens when a payment goes off the rails, because an agent has done something unauthorized, and do you see any problems there that Congress should be addressing? The other one is just facing the problem with the super apps. I mean, we probably all saw a couple of days ago, Facebook just bought a financial super app with 3 billion customers, like 10 times the population of the United States, that does everything: that does payments, it does loans, it does business loans, the whole thing. And so I’m very worried that we are going to be facing this sort of vertical monopoly in the United States in a way that will really crush competition. Let me just go down the line and sort of tell me your thoughts briefly on those two issues.”

Paige Pidano Paridon, BPI: “Sure. On agentic payments, I’ll start with that. I agree. I think the advent of that technology and capability raises a lot of questions that are just beginning to be investigated and considered, and I do think regulators and the Congress should carefully consider all of the possible implications, including, as you note, what happens if an agent executes a transaction or makes a payment that it wasn’t authorized to do, and who is a consumer in that situation? Who is the institution serving, and in the investment advisor space, who are the fiduciary duties owed to, etc.”