Agentic Commerce Without the Panic: A Practical Playbook for Fraud and Risk | Presented by Signifyd
Summary
Agentic commerce is coming, we know. But for fraud and risk teams, the real question isn't whether to embrace it, it's how to enable it safely without getting ahead of your fraud controls, your liability framework or your ability to distinguish a legitimate AI-initiated order from an abusive one.
Agentic transactions introduce a fundamental detection problem. The behavioral and device signals fraud models depend on were built for human shoppers. When the buyer is an AI agent, even a legitimate one acting on behalf of a real customer, much of that signal degrades or disappears. Chargeback liability gets more complicated too: when a consumer authorizes an AI agent to purchase on their behalf and later disputes the transaction, existing frameworks weren't designed to answer the question of who owns it.
Meanwhile, consumer advocate agents, AI systems that automatically file refund requests, escalate warranty claims and negotiate price adjustments on behalf of customers, are beginning to surface in merchant queues, creating a new category of post-purchase abuse risk that most teams aren't yet equipped to handle.
In this session, Signifyd is joined by fraud and payments leaders from Wayfair and Pandora, two major retailers at the leading edge of agentic commerce preparation, for an honest, practitioner-level conversation about what they're building now, what they're watching and what they'd prioritize if starting from scratch.
Topics will include detection strategy in a low-signal environment, emerging identity and credentialing standards such as Visa's Trusted Agent Protocol and Google's Agent Payments Protocol, chargeback liability and policy design and how to get fraud and risk infrastructure into the agentic commerce conversation before the architecture decisions are already made.
What You'll Learn
- A practical framework for understanding the spectrum of agentic commerce interactions and the distinct fraud and risk profile each one presents.
- A clear-eyed view of the signal blackout problem and the detection approaches most likely to hold up when traditional behavioral signals degrade.
- Insight into how two major retailers are approaching agentic commerce preparation today, including where they are investing now and what they are waiting on.
- A working understanding of the policy and tooling implications of consumer advocate agents and how to prepare for post-purchase abuse before the volume arrives.
- Concrete guidance on how fraud and risk leaders can engage earlier in agentic commerce planning to ensure trust infrastructure is built in from the start.
Speaker
