Building Trustful and Fraud-Resilient AI: The Critical Role of Training Data

AI is already making high-stakes decisions across moderation, fraud, AML, and platform integrity. But when training data fails, the fallout isn’t theoretical, it’s immediate. Scams get through, bad actors scale, bias deepens, and regulators start paying attention.

In this high-impact session, we’ll unpack how model training has become a frontline control for risk, safety, and compliance. We’ll explore how to identify early signs of model weaknesses and how to build resilient AI tools to support platform safety and resistance to threats.

Joined by platform leaders sharing real-world experience, we’ll go beyond theory to show how organizations are tackling these challenges at scale.