Webinar Description
Key Takeaways
- Examines why expanding fraud detection signals has not consistently improved outcomes for financial institutions
- Explores the role of agentic AI in fraud decision-making and the growing importance of explainability
- Addresses signal quality, durability and contextual relevance over raw data volume
- Relevant for fraud, digital and risk leaders in financial services
Introduction
A webinar scheduled for August 2026 will examine a growing paradox in financial services fraud prevention: despite significant investment in expanding data sources and detection signals, many institutions have not seen proportional improvements in fraud outcomes. The session brings together industry thought leaders to discuss why signal quality and decision intelligence may matter more than data volume in the evolving fraud landscape.
The Data Accumulation Problem
Over the past decade, financial institutions have dramatically expanded the range of signals available for fraud detection. Device intelligence captures information about the hardware and software used in transactions. Behavioural intelligence analyses patterns in how users interact with digital services. Identity data, transaction patterns, consortium data shared across institutions, and biometric verification have all become standard components of fraud prevention stacks.
Yet this accumulation of data sources has not automatically translated into better fraud prevention. The challenge lies not in the availability of signals but in their effective integration and interpretation. Raw data volume can introduce noise, create false positives, and complicate decision-making processes rather than streamlining them.
Signal Quality and Durability
One of the central themes of the discussion concerns the distinction between signals that provide lasting value and those that quickly lose effectiveness. Fraudsters continuously adapt their techniques, meaning that signals which prove highly predictive today may become unreliable as attack methods evolve. Financial institutions face the ongoing challenge of identifying which data sources offer durable insights and which represent diminishing returns on investment.
Context also plays a critical role. A signal that proves valuable in one transaction scenario may be irrelevant or misleading in another. The ability to apply appropriate context to data interpretation separates effective fraud prevention from systems that generate excessive false positives or miss sophisticated attacks.
Agentic AI and the Explainability Imperative
As financial institutions invest in artificial intelligence and next-generation technologies, the discussion will address how agentic AI is reshaping fraud decisions. Agentic AI systems can autonomously execute complex workflows and make decisions with minimal human intervention, offering potential efficiency gains in fraud detection and response.
However, this autonomy raises significant questions about explainability. Regulatory requirements and internal governance standards increasingly demand that institutions can explain why specific decisions were made. When AI systems operate with greater independence, the ability to trace decision logic back to underlying signals becomes essential for compliance, customer relations and continuous improvement of fraud models.
Who Should Attend
The session is designed for professionals responsible for fraud prevention strategy, digital transformation and risk management within financial institutions. Those evaluating AI investments for fraud detection or seeking to optimise existing signal infrastructure will find the discussion particularly relevant. The themes addressed reflect challenges common across banking, payments and broader financial services.