Trust & Safety leaders know that preventing harm is the right thing to do—but budget owners want to know what it is worth. How do you credibly show ROI when the “return” is an incident that never happens, a victim who is never harmed, or a headline that never hits? This panel will tackle the uncomfortable but essential challenge of putting a price on harm in order to secure investment in safety. Panelists will share practical methods for estimating the cost of incidents—from fraud losses and support costs to reputational damage, regulatory exposure, and long-term user trust. They will discuss how to frame victim impact ethically and respectfully while still translating it into the financial language of CFOs, GMs, and boards. Through real examples and candid discussion, the session will cover how to build defensible cost models, use scenarios and ranges instead of false precision, and construct narratives that connect safety outcomes to growth, retention, and risk mitigation. Attendees will leave with actionable tools to move beyond “it’s the right thing to do” and effectively sell ROI for Trust & Safety initiatives without diminishing the human impact at the heart of their work.

As digital platforms scale AI-driven Trust & Safety, automation often introduces new operational risks such as model drift, brittle decisioning, and rising human override costs. This session explores how to design AI-assisted risk systems that scale globally while remaining reliable, transparent, and defensible. Drawing on real-world experience operating large, multi-market platforms, the talk focuses on practical design choices: where automation creates durable value, where it breaks down, and how to build resilient human-in-the-loop workflows. Attendees will leave with actionable frameworks for balancing automation, operational stability, and regulatory expectations at scale.

Fraud is never just “out there.” For modern marketplaces, risk lives everywhere: in internal tools, in outsourced operations, and at the edges of the platform where sophisticated external actors test every control. At the same time, vendors and partners operate under commercial pressure, staff turnover, and incentive structures that can make them vulnerable to mistakes—or, in rare cases, misconduct and corruption. This panel will explore fraud as an end-to-end ecosystem problem, not just a point solution or a “T&S issue.” We will examine how fraud surfaces across three layers: In-house: employees with privileged access, policy discretion, and deep system knowledge. Outsourced / Vendors: BPOs and third parties who handle reviews, support, or operations, often with high access and lower visibility. External Bad Actors: organized fraud rings, opportunistic users, and third-party toolkits constantly probing for weaknesses.