Financial institutions have multiple obligations to surveil the activity and behavior of their clients, counterparties, and associates. The scope of these obligations is growing as a higher percentage of client interactions are digitized, creating the expectation by the regulator that everything that can be surveilled should be. Additionally, regulators are increasingly deploying their own tools to monitor the data submitted by institutions and looking for patterns where the institution should have acted but didn’t. Lastly, internal and external bad actors are becoming more sophisticated in using digital tools and capabilities to commit fraud or other market-manipulating activities.
In response to the above-listed issues and challenges, most financial institutions have invested heavily in technology solutions that improve their ability to ingest the appropriate data and identify potentially illegal behaviors. Historically, the infrastructure used to perform surveillance (people, process, and technology) has been built in silos that are dedicated to the specific surveillance obligation. Financial Crime Compliance requires monitoring of transactions for suspicious activities and potential fraud, market abuse regulations require trading activity to be surveilled for potential manipulation, and communications with clients need to be monitored to detect unethical or criminal conduct. The surveillance methods across these scope areas are often different because some activities, like fraud detection, are real-time, and some activities, like AML transaction monitoring, are done T+1 and often in batch mode. The data used in this monitoring activity overlapped but was different, driving the bespoke infrastructure.
As data management and analytic technologies have advanced, including big data, machine learning, natural language processing, and generative AI, the underlying architecture across the different surveillance solutions is converging. This creates a consolidation opportunity, allowing institutions to rearchitect their infrastructure, maximize common components, and reuse using a combination of in-house developed and vendor solutions. Each layer in the stack can be optimized and shared across the different obligations and business units, including data ingestion, data prep, pattern detection, alert generation and disposition, and case management.
There are tremendous benefits from a consolidated surveillance infrastructure, including lower cost of compliance, increased scope of covered activity, more accurate and effective alerts (fewer false positives and more true positives), and better utilization of the operational and analyst resources.
Achieving these benefits requires a different approach to your surveillance technology architecture, which I will explore in subsequent posts.
