Banks struggle to turn AI surveillance spending into results

New research from 1LoD, commissioned by Shield, says banks are increasing AI investment in communications surveillance but still face the same core problems seen in 2024. The report points to a widening gap between AI ambition and operational reality, with false positives, fragmented data and legacy systems still blocking progress.

Why it matters: - Banks are spending more on AI-driven surveillance, but the report says many firms are not seeing the operational gains needed to improve risk detection at scale. - More than 90% of institutions still operate in alert-heavy environments, which keeps review teams buried in noise and slows down real escalation. - The findings suggest the main obstacle is now execution, not regulation, which changes where banks need to focus investment and modernization.

What happened: - 1LoD published a new report, "The Future of Surveillance: Turning AI Ambition into Operational Reality," based on responses from senior surveillance and compliance professionals at financial institutions. - Shield commissioned the research and said the report compares current findings with 1LoD's 2024 benchmarking survey. - The report was released Sept. 9, 2026, in New York.

The details: - False positives were the top-rated surveillance challenge, with 52% of firms calling them highly significant. - Budget and resourcing constraints were cited by 41% of respondents. - Limited access to quality data and legacy or outdated systems were each cited by 37% of firms. - Only 7% of respondents rated regulatory pressure as highly significant. - Shield said deployment data from evaluations spanning millions of communications showed roughly a threefold reduction in alert noise compared with legacy systems. - In direct comparisons, Shield said the platform delivered up to 44% higher accuracy and about three times more actionable escalations. - Shield's platform covers data ingestion, regulatory-grade archiving, discovery and investigations, surveillance, voice intelligence and agentic AI capabilities. - The platform uses a unified, cloud-based data foundation intended to create a single source of truth for communications governance.

Between the lines: - The report frames the industry's problem as an execution gap: AI tools are becoming more available, but banks are still constrained by fragmented data, legacy infrastructure and operating models built for a different era. - Leading compliance teams are moving from static, rules-based detection to adaptive models that learn from behavior, while also unifying data and workflows into integrated platforms. - The report argues that firms tying surveillance spending to measurable cuts in manual review are more likely to turn AI investment into better detection outcomes. - Shield's results are presented as an example of how reducing alert noise can increase the share of escalations that matter.

What's next: - The report expects surveillance to keep moving toward integrated data, context-driven detection, reduced operational noise and measurable outcomes. - Institutions that close the gap between available technology and day-to-day practice could lower costs while improving true risk detection. - The full report is available online.

The bottom line: - Banks are not short on AI ambition; they are short on the operational plumbing needed to turn that ambition into better surveillance.