Survey Sees Compliance “Alert Fatigue”

Compliance teams are increasingly being pressured by rising “false positive” alerts, raising the number of manual interactions and creating “alert fatigue”, according to a new report.

The report, by Eventus and Datos Insights, finds that our 70% of respondents from 20 banks and broker-dealers, had false positive rates above 25%. This means the report observes, already stretched compliance teams add to their manual workload as well as an increased risk of missed threats that can cause important signals to be overlooked.

While the majority of firms interviewed have adopted cloud-based or hybrid technology platforms, many still struggle with siloed data, legacy systems, and inconsistent integration across asset classes and regions, the firms reveal. “Addressing automation with smarter calibration and better data controls is a high priority,” they state.

As is the case across most business segments, the volume and quality of data are a key challenge as surveillance depends on accurate, complete data, the report states. Trade surveillance systems must process vast amounts of data in real time from multiple sources – including order management systems, market data feeds and execution records.

The data, often complex, comes in various formats, making timely processing and analysis a significant challenge. The report highlights how inaccurate timestamps, misaligned or missing data and partial execution records can undermine the effectiveness of surveillance. “Ensuring clean and complete data is critical for accurate detection of suspicious trading patterns, yet it remains a persistent issue,” the report warns.

Inevitably, the report sees greater use of AI, noting that firms are shifting from rules-based surveillance to data-driven frameworks powered by AI, with adaptive filtering, real-time analysis and integrated case management seen as a key opportunity to reduce alert fatigue. As AI adoption accelerates, firms must maintain control and transparency, it warns, adding whether this is achieved through tuning thresholds or auto-closing alerts, workflows must remain auditable and defensible – particularly in the eyes of regulators.

Market participants are focused on development frameworks that support surveillance across equities, derivatives, fixed income and digital assets. In particular, integrating the latter two asset classes with the wider surveillance function can be challenging due to fragmentation or low liquidity in certain names, the report says. While firms aim to establish enterprise-wide integrated surveillance with strong governance, practical implementation is often segmented by asset class and legal entities due to jurisdictional requirements and operational complexities.

This fragmented approach results in varied surveillance models across regions, requiring tailored compliance measures while maintaining overarching governance standards, the report observes. Further, firms are challenged with keeping systems up to date with the latest legal requirements, as well as adapting models to new trading strategies, such as for digital assets or new asset classes.

The report also finds that institutions are re-evaluating build-vs-buy strategies as customisation demands increase, while many are supplementing vendor platforms with in-house tools. Equally, the surveillance workforce is evolving to include data scientists, behavioural analysts and ex-traders, reflecting the growing sophistication, complexity and scale of surveillance operations.

“This report surfaces both the challenges and exciting innovation taking place in surveillance functions around the world,” says Travis Schwab, CEO of Eventus. “From deploying AI and advanced analytics to tackling persistent market and data fragmentation, firms are making important strides but still facing significant obstacles.”

Vinod Jain, strategic advisor at Datos Insights, adds, “Effective market surveillance is not just about monitoring; it’s about precision, adaptability and leveraging technology to turn high-quality data into actionable insights. As the industry faces growing complexities, integrating AI-driven models and unified frameworks will be key to reducing false positives, strengthening data integrity, and ensuring compliance across jurisdictions.

“Without reliable, standardised data, even the most advanced surveillance systems risk inefficiency and missed opportunities for detection,” he continues. “False positives are an inherent cost of surveillance, but refinement is key. A balanced approach – where automation enhances efficiency while human validation ensures accuracy – creates a system that is both intelligent and reliable.”