当犯罪分子以机器速度行动时，金融犯罪运营模式必须适应：管理服务的必要性

The case for managed services The cost of stayingstuck in the past Financial institutions are investing more infinancial crime (FinCrime) compliance than atany point in history, yet they continue to fallfurther behind. Criminals innovate faster thanbanks can respond; real-time payment rails,the rise of digital assets and the emergenceof AI-driven threats have amplified bothfraud and money laundering risks. Globalenforcement actions remain unrelenting. More effort ≠better outcomes $10b AML penalties globally 2025 In 2025 alone, anti-money laundering(AML) penalties exceeded $10 billion,with US regulators driving much of theenforcement.1At the same time, traditionalalert-centric operating models are failing,delivering up to 95% false-positive rates,2overwhelming investigative teams anddiverting the expertise required to identifytrue risk. The shift to real-time paymentshas raised the stakes further Volumes and values on networks such as RTP and FedNowhave surged,3while authorized push payment fraud, muleactivity and synthetic identity schemes scale at machinespeed. Legacy batch-processed FinCrime programs simplycannot match this velocity. Meanwhile, the human capitalmodel underpinning most in-house programs is showingsigns of stress: Repetitive, high-volume, Level 1 (L1) workdrives chronic turnover, leaving persistent capacity gapsand degraded control quality.4 at the exact moment when criminals scale and real-timepayments eliminate any margin for delay or error. For financial services leaders, the choice is no longerbetween outsourcing and building in-house; it is betweena model that can scale with modern FinCrime and onethat cannot. The institutions that move first to a managedservices model will reclaim investigative capacity, reducerisk, meet rising regulatory expectations and build resiliencefor the next wave of real-time payment expansion andgeopolitical volatility. Those that wait will face rising costs,widening gaps and increasing enforcement exposure. The conclusion is unavoidable: Running FinCrime operationsentirely in-house has become a strategic liability. It yieldsrising costs, inconsistent outcomes and operational fragility Today Near future Beyond Scaled managed services thatembed cognitive investigatorsand automate qualityassurance and reportingthrough agentic processes Autonomous compliance ecosystemsthat are orchestrated using AImarketplaces, predictive risksensing, geopolitical scenarios andcontinuous regulatory assurance Pilot AI-led alert triage,vendor-managed investigations,data quality uplift, governanceand regulator engagement 3“RTP’s Instant Payment Volume Nearly Doubled in 2024,”PaymentsJournal website, www.paymentsjournal.com/rtps-instant-payment-volume-nearly-doubled-in-2024/, January 13, 2025. 4“How to increase financial crime analyst capacity without increasing headcount or burnout,”WorkFusion website,www.workfusion.com/blog/how-to-increase-financial-crime-analyst-capacity-without-increasing-headcount-or-burnout/, August 24, 2023. The harsh reality:in-house models areno longer sustainable While institutions continue to expand their FinCrimebudgets, the operational foundations of in-house programsare weakening in ways that incremental investment cannotfix. The core issue is not simply rising volume or regulatorypressure: The traditional, internally staffed model isstructurally mismatched to the speed, scale and complexityof today’s FinCrime landscape. Despite sizable compliance spend, even global banks withsignificant resources continue to fall short of regulatoryexpectations. Record high settlements in recent yearsunderscores that conventional internal programs, evenwhen mature and well resourced, are struggling todemonstrate sustained effectiveness under modernsupervisory scrutiny. This reflects a broader trend:Investments increasingly flow into labor-intensive processesrather than driving meaningful improvements in detectionor control strength, creating a trajectory where programsgrow costlier without becoming more capable. With ashortage of qualified talent increasing the cost of labor, fullystaffing an operation is turning into a challenge that can’t beconquered by any but the largest institutions. Outcomes A major driver of this inefficiency is the fragility ofthe human capital model underpinning most in-houseoperations. Analysts shoulder repetitive, high-volumeinvestigative work, driven largely by noise rather than risk.As a result, predictable L1 attrition near the 12-monthmark creates chronic gaps and forces institutions intocontinuous cycles of hiring and retraining.5This instabilityis compounded by industry-wide monthly separation rates of 2% to 3%, which disproportionately affect specialized functions that dependon retained expertise and continuity. The end state is a fixed-capacity operationthat cannot flex when payment volumes surge, typologies evolve or regulatoryexpectations tighten. Complicating this further is the fact that traditional ale