Marketplaces are stepping into a new risk era where traditional fraud models collapse under the speed and sophistication of AI driven attacks. Deepfake identities, AI generated merchant profiles, synthetic seller networks, and hyper personalized social engineering are no longer theoretical they are scaling faster than most platforms can detect. This session will deconstruct how AI is reshaping fraud behavior, what this means for global marketplaces, and why legacy trust and safety frameworks are structurally unprepared. The content will be grounded in practical, operational insights for marketplace founders, risk leaders, and compliance executives who need to modernize their defenses without slowing growth.

As companies integrate autonomous AI, they face a new “agentic” attack surface where agents can bypass traditional security controls and trigger cascading risks, from account takeovers to identity manipulation. This session examines real-world threats, the evolving regulatory landscape—including California’s transparency mandates and the EU AI Act’s downstream liabilities—and why static defenses are insufficient against autonomous systems. Attendees will explore how unplanned automation creates hidden risk and gain a practical roadmap for shifting from simple human oversight to resilient, objective-based governance that secures both technology and its users.

Early-stage and/or high-growth companies face fraud risks long before they have the budget, tooling, or headcount of mature platforms. This panel explores pragmatic, resource-conscious approaches to fraud prevention, detection, and response, drawing on real-world experience from companies scaling rapidly under operational constraints. Panelists will discuss how to prioritize risks, make build-versus-buy decisions, and evolve fraud programs responsibly as companies grow.

Every platform runs investigations. Almost none retain what those investigations produce. The findings go into reports, Slack threads, and the heads of analysts who eventually leave. The next team that hits the same pattern starts from scratch. This talk draws on 12 years of running investigations at platform scale. I'll walk through how investigation knowledge disappears, what that costs operationally, and what a system designed to retain and compound investigative learning would actually look like. Not a new dashboard. Infrastructure that governs what happens after the report.

Digital marketplaces are facing increasingly sophisticated fraud, from coordinated identity networks to promo abuse and account takeovers. Traditional rule-based systems are struggling to keep pace, often increasing false positives and customer friction. This session explores how marketplaces can transition to AI-native risk systems powered by machine learning, graph analytics, and real-time decisioning. Drawing on LatentView’s experience, we will discuss building unified trust scores, enabling real-time fraud detection, and augmenting fraud operations with GenAI copilots. Attendees will gain practical insights on balancing fraud prevention with customer experience while moving toward scalable, continuously learning risk and trust ecosystems.