Content will be focused on how AI and machine learning models can be used for scaling living room safe and child safe content. With the amount of content that is existing it is imperative there is some age gating that needs to happen but it is not scalable through manual reviews. AI and ML models can help achieve that.

What do “free iPhones,” “free food,” and “free stays” have in common? They’re all outcomes of today's GenAI-fueled fraud ecosystem, where industrial-scale tactics are used to bypass security, drain merchant incentives, and exploit identity gaps. AI isn’t just transforming platforms. It’s powering their attackers. As generative AI and virtual environments become more accessible, fraud tactics are evolving from manual exploits to industrialized, automated operations. From emulators and app cloners to deepfake-powered account creation, the new fraud playbook is faster, more scalable, and harder to detect. Fraudsters are becoming increasingly organized, resembling a production line, and these tools are being distributed on the deep web and in fraudster communities on social networks and messaging apps, operating as a monthly subscription service costing a few hundred dollars: it is fraud-as-a-service. In this session, we'll unpack how AI is being weaponized in today’s fraud ecosystem and what real-world defenses look like. Expect practical case studies, data-driven insights, and a breakdown of the signals that still surface truth in a landscape of synthetic noise. Key Takeaways: - How generative AI and automation are fueling large-scale fraud schemes - How device, location, and behavioral signals expose hard-to-catch threats like deepfakes and virtual farms - Real-world examples of how fraud leaders are adapting their stacks to stay ahead of evolving threats

Marketplaces are seeing rapid growth in wellness aids, performance enhancers, research chemicals, and other health-adjacent products that sit in regulatory gray areas. Listings change quickly and sellers adapt faster than policy. The consequences of missing the mark on this type of risk (including consumer harm, enforcement action, and brand damage) can be immediate — and significant. This session covers emerging wellness risk areas and what they mean for marketplace policy, enforcement, and reputational exposure in 2026. Drawing on real-world patterns, we examine how these products appear on platforms, why they escalate risk, and how marketplaces can prepare to meet the challenges. Attendees will learn to: - Understand the risk profiles of key categories, including weight-loss drugs, novel cannabinoids, kratom alkaloids, peptides, and ingredients subject to country-specific restrictions. - Recognize seller tactics that complicate detection, such as “research use only” positioning, euphemisms, bundling, and jurisdictional arbitrage. - Identify practical policy and enforcement implications, including where to tighten category definitions, when enhanced review is warranted, and what signals to monitor to stay ahead of regulatory attention.

Drift is often treated as a system failure, but in practice it is an early signal of misalignment. In digital marketplaces, models can remain accurate while trust erodes, fraud adapts, or safety outcomes worsen. This session reframes drift as the measurable gap between system outputs and how humans interpret, rely on, or delegate those outputs over time. Using real marketplace scenarios, I introduce a simple Goals–Rules–Thresholds (GRT) lens to detect drift early, distinguish healthy adaptation from rising risk, and define clear intervention points—before drift becomes harm, regulatory exposure, or public failure.

Marketplace Trust & Safety, fraud, and compliance teams increasingly rely on AI to make high-impact decisions affecting user access, livelihoods, and regulatory exposure. Yet many governance failures are not model failures, but accountability failures—automation obscures ownership, rationale, and uncertainty. This session reframes AI governance as judgment infrastructure, not automated enforcement. Drawing on experience building large-scale marketplace integrity systems, the talk shows how defensible governance emerges from workflow design: where AI should surface risk, where automation must stop, and which decisions require explicit human ownership. Attendees will leave with a practical framework for designing AI systems regulators, operators, and marketplaces can defend.