Navigating the Data Labyrinth: How Major Banks Are Tackling Risk Reporting, Legacy Infrastructure, and AI Integration

The modern banking sector finds itself at a critical juncture, wrestling with the dual pressures of escalating regulatory demands and the imperative to modernize legacy infrastructure. Fragmented data architectures and recurring manual interventions continue to severely undermine institutional risk reporting capabilities. According to recent supervisory insights from the European Central Bank (ECB) released in February 2025, persistent data control weaknesses leave major financial institutions acutely vulnerable, hampering their capacity to respond swiftly and decisively during periods of systemic market stress.

These regulatory warnings are not isolated observations but part of a broader, systemic trend. A comprehensive thematic review published by the Basel Committee on Banking Supervision (BCBS) in November 2023—utilizing baseline data compiled as of June 2022—underscored the staggering scale of the challenge. The assessment revealed that out of 31 global systemically important banks (G-SIBs) evaluated worldwide, a mere two institutions were fully compliant with all 11 principles for effective risk data aggregation and reporting, known commonly as BCBS 239. This widespread compliance gap has intensified scrutiny from international watchdogs, forcing financial institutions to rethink how they govern enterprise data, manage risk, and deploy emerging technologies such as artificial intelligence (AI).

Against this backdrop of heightened regulatory scrutiny, the operational realities of scaling automation across complex, siloed banking ecosystems have come under intense examination. Nathaniel Bell, Data Management Director at Wells Fargo, recently joined the Emerj AI in Business Podcast to discuss these exact frictions. In conversation with host Yolandi de Weerdt, Bell explored how inconsistent data interpretations, execution evidence gaps, and mounting oversight requirements shape executive decision-making. Specifically, the dialogue addressed how banking leaders evaluate AI investment returns and navigate the operational hazards of scaling technology atop unstable foundations.

The Regulatory Chronology and Escalating Compliance Pressures

To understand the current urgency surrounding data modernization, one must examine the timeline of regulatory milestones that have shaped the risk reporting landscape. The journey toward rigorous data aggregation standards began in the wake of the 2007–2008 global financial crisis, when regulators recognized that institutions lacked the internal visibility required to assess their aggregate risk profiles in real time.

In response, the Basel Committee issued its foundational BCBS 239 standard in January 2013, establishing a strict deadline for global systemically important banks to achieve full compliance by early 2016. However, compliance proved exceedingly difficult. Subsequent progress reports issued by the Basel Committee throughout the late 2010s and early 2020s repeatedly highlighted sluggish implementation speeds, data siloing, and a heavy reliance on manual spreadsheets.

By the time the November 2023 BCBS compliance report was published, revealing that only two out of 31 G-SIBs had achieved full compliance, it became glaringly evident that traditional remediation strategies were falling short. This historical friction was further compounded in February 2025, when the ECB issued its supervisory findings, explicitly tying ongoing risk reporting deficiencies to structural weaknesses in legacy data controls. As regulatory bodies worldwide shift from consultative guidance to active enforcement and potential capital add-ons for non-compliant institutions, banking executives face mounting pressure to accelerate structural reforms.

Standardize Workflows Before Scaling Automation

Large financial institutions rarely operate on pristine, unified technology stacks. Through decades of mergers, acquisitions, product line expansions, and shifting regulatory mandates, they accumulate a mosaic of legacy platforms, disparate databases, and fragmented workflows. Within this environment, individual operating units often operate under the assumption that they are executing processes correctly within their local software silos. Yet, the enterprise as a whole frequently struggles to generate consistent, reliable outcomes across the complete end-to-end workflow.

As Bell noted during his discussion, the underlying problem encompasses both data access and shared organizational context. A routine business process—such as commercial loan origination, anti-money laundering (AML) screening, or customer document verification—frequently traverses multiple software systems and cross-functional teams without establishing a unified, standardized interpretation of the data being handed off. Furthermore, evolving regulatory mandates introduce additional layers of operational complexity, requiring institutions to translate dense, written legal guidance into consistent operational workflows.

When financial institutions introduce artificial intelligence or advanced automation tools into unstandardized environments, they risk amplifying these existing flaws. Rather than solving structural inefficiencies, automation can rapidly scale errors, a phenomenon Bell describes as “industrializing bad data quality.”

To mitigate this risk, banking leaders must prioritize end-to-end workflow standardization before committing capital to AI deployment. According to Bell, executives should analyze the complete lifecycle of a process to determine whether a proposed technological tool genuinely accelerates an effective workflow or merely papers over systemic inconsistencies.

For instance, consider a routine customer-document review process where two separate departments apply conflicting validity periods to the same compliance certificate. Introducing an automated routing tool will undeniably speed up the physical transfer of the digital document from the first team to the second. However, the receiving team remains burdened with resolving the underlying interpretative conflict. By resolving the ambiguity at the input stage through standardized acceptance rules, the institution fundamentally alters—and often simplifies—the work required to complete the review.

Bridging Control Gaps and Reducing the Manual Verification Burden

Technology vendor demonstrations frequently showcase impressive software capabilities without addressing how an institution will verify, audit, and monitor the tool’s performance in a live production environment. Bell emphasizes that technology investments must be anchored in clear business objectives, ensuring that operational outcomes can be measured, audited, and continuously refined.

Establishing verifiable oversight is a formidable challenge in modern banking. While process maps outline the intended sequence of operational actions, reconstructing the actual sequence of events becomes increasingly difficult as data flows across isolated legacy systems. Bell identifies this challenge as reconciling "what actually happened" against the documented expectations of the workflow.

To bridge this control gap, institutions can leverage automated system logs and immutable audit trails to demonstrate that mandatory operational steps occurred and to flag exceptions in real time. For example, if bank policy dictates that customer onboarding documentation must undergo rigorous compliance review before an application is approved, separate legacy systems may require staff to manually harvest records to prove the review took place. By capturing structured execution metadata at each phase, compliance teams can automatically verify whether applications advanced strictly in accordance with policy.

However, capturing and maintaining this operational evidence creates secondary labor burdens. Bell characterizes the traditional generation of audit and compliance evidence as a “heavy tax” on human resources. When evaluating an AI or automation investment, banking leaders must calculate whether the technology actively reduces this verification tax or simply shifts the administrative burden elsewhere. A workflow optimization that accelerates transaction processing is ultimately incomplete if compliance staff must still spend hours manually reconstructing audit trails after the fact.

Evolving Governance: Reducing Manual Oversight While Preserving Accountability

As regulatory expectations expand, financial institutions face a continuous influx of new compliance controls, audit demands, and review cycles. Meeting these growing obligations by simply hiring more personnel or adding manual oversight layers is unsustainable. Instead, modern governance frameworks must embed controls directly into core operational workflows and the underlying data products that support them.

This structural shift transforms the daily responsibilities of risk and governance professionals. Rather than spending valuable time manually reviewing dense documentation, coordinating cross-departmental sign-offs, and reconciling conflicting spreadsheets, governance teams can pivot toward monitoring systemic signals and investigating structural anomalies.

In this modernized operating model, automated monitoring systems detect behavioral patterns that deviate from historical baselines, signaling potential operational failures or emerging risks. Interpreting these signals, however, requires deep institutional knowledge encompassing metadata structures, data lineage, operational controls, and ultimate business outcomes.

Bell draws an insightful parallel between adopting advanced enterprise technology and purchasing home exercise equipment: the ultimate value derived from the investment depends entirely on sustained behavioral changes and consistent usage patterns. Similarly, financial institutions must actively reshape how human teams collaborate with technology to achieve intended business outcomes.

If a banking unit introduces automated document processing but employees continue maintaining parallel, manual tracking spreadsheets out of habit, the institution fails to capture efficiency gains. Leadership must actively determine whether new automated records fully satisfy internal and external evidentiary standards, thereby allowing legacy manual tasks to be safely retired.

Strategic Implications for Banking Executives

The insights shared by Nathaniel Bell offer a pragmatic framework for banking executives navigating the complex intersection of data management, regulatory compliance, and artificial intelligence integration.

1. Prioritize Workflow Clarity: Before deploying capital toward AI or advanced automation, institutions must diagnose and standardize end-to-end workflows to eliminate data ambiguity and prevent the industrialization of poor data quality.
2. Automate Evidence Capture: Technology investments must be evaluated based on their ability to reduce the "heavy tax" of manual audit preparation by embedding verifiable controls directly into transaction execution logs.
3. Redefine Governance Roles: Compliance and risk professionals should transition from manual document checkers to strategic investigators who monitor system-generated risk signals and manage operational exceptions.

By aligning technology deployment with rigorous data standardization, robust execution verification, and intelligent oversight, financial institutions can move beyond reactive compliance. Ultimately, banks that successfully address these foundational data challenges will be better positioned to harness artificial intelligence, streamline operations, and withstand the rigorous demands of modern banking supervision.