The Expert View: AI and the Reinvention of Credit Risk in Commercial Banking

Across the banking industry, a fundamental shift is underway. Artificial intelligence is no longer an experimental add-on or a distant ambition—it is rapidly becoming the core engine reshaping how credit risk is assessed, how decisions are prepared, and how value is ultimately delivered to customers.

 

“We’re going through a fundamental shift that could impact every part of banking operations,” said Anthony Morris, Chief Industry Innovation Officer at nCino, speaking to executives at a Business Reporter Breakfast Briefing in Amsterdam. “It’s an operating-model change. For the first time, the activation of data is not just compressing processes, but removing large amounts of administrative overhead and enabling better insights and decisions. This is a tectonic shift.”

 

While banks are advancing at different speeds, few doubt the direction of travel. Artificial intelligence—particularly generative AI—is steadily becoming embedded across the entire credit risk lifecycle, from origination through to portfolio monitoring and ongoing management. 

 

As Niels Geneste, Regional Vice President at nCino, noted: “Banks are increasingly looking to consolidate multiple technology islands onto a single platform, with the goal of automating as much of the process as possible.”

 

A way forward to data-driven decisions

 

Generative AI offers a way to combine internal customer data, peer and sector benchmarks, and external macroeconomic indicators, AI systems can interpret policy dynamically rather than simply applying it mechanically.

When full credit policies are embedded into large language model-based systems, banks can assess new credit applications and existing exposures in near real time. These systems can also benchmark applicants against comparable entities and provide transparent explanations for decisions, grounded in specific policy clauses and underlying data sources.

 

The purpose of AI is not to replace human judgment. Its real contribution lies in removing the manual burden of collecting, validating, and reconciling information—allowing risk professionals to focus on interpretation, challenge, and ultimately higher-quality decision-making.

 

Where AI is delivering value today

 

In practice, AI agents are already proving most effective when applied to clearly defined components of the credit process. Financial spreading, document analysis, covenant tracking, and early-warning signal detection are all areas where automation is delivering measurable gains, while human experts retain responsibility for validation and final approval.

 

This modular approach is also critical for transparency and accountability. Allowing AI to operate as a “black box” across the entire process introduces unacceptable risk in regulated environments. Instead, breaking workflows into discrete, auditable steps ensures that institutions can demonstrate how conclusions were reached—what data was used, which assumptions were made, and how policies were applied.

 

There is also strong momentum behind extracting structured insights from unstructured data such as contracts, financial statements, and supporting documentation. These capabilities are already well within reach.

 

Redefining risk management roles

 

The adoption of AI will also require a broader organisational transformation. AI is clearly absorbing entry‑level analytical work—spreading, document review, policy checks—the very tasks through which junior bankers traditionally learned the craft. If learning by repetition disappears, banks must consciously redesign how judgment is taught. At the same time, a counter‑argument challenged nostalgia: handwriting disappeared too, yet literacy survived. The real risk is not loss of old skills, but failure to define new ones.

 

Some banks will become more transaction-oriented and accelerated, while others will shift toward advisory functions, scenario analysis, and the translation of complex data into client-facing insight.

 

Data as a strategic differentiator

 

Data remains the unsolved bottleneck. While AI promises richer insights, its impact is uneven. Large corporates generate abundant public data; SMEs do not. ESG is automatable for multinationals but still questionnaire‑driven for small firms. This creates a paradox where AI risks widening informational inequality across customer segments unless banks collaborate on shared utilities or anonymized data pools. However, the data ecosystem is maturing quickly. New entrants are investing heavily in aggregating, cleaning, and enriching large-scale datasets, while banks continue to hold valuable proprietary client information.The combination of curated external datasets and internal data assets creates a powerful foundation for more granular insight and more tailored client engagement.

 

Governance, accountability, and regulation

 

Across the discussion, governance emerged as one of the most significant unresolved challenges. Many organisations still lack clear frameworks for how AI systems should be owned, monitored, and challenged. A fragmented, bottom-up experimentation approach risks creating inconsistent controls and unmanaged exposure.

 

What is required instead is a coordinated, top-down strategy that integrates technology adoption with cultural change, skill transformation, and operating model redesign.

 

Regulation adds another layer of complexity. Banks remain fully accountable for decisions, even when those decisions are heavily influenced or prepared by AI systems. In some cases, regulatory uncertainty has already constrained previously effective operating models, particularly in areas involving automated decisioning.

 

While regulation will inevitably lag behind technological innovation, there is broad consensus that scrutiny will intensify as AI becomes more deeply embedded in core banking functions.

 

This raises a fundamental question: as decision throughput increases, can “human-in-the-loop” remain a meaningful safeguard, or does it risk becoming a procedural formality? If humans are expected to approve more decisions, faster, and with less visibility into underlying logic, then accountability structures must be explicitly designed—not implicitly assumed.

 

Looking ahead

 

AI will not eliminate risk from banking, but it will fundamentally redistribute it. Less risk will sit in data collection and processing; more will concentrate in judgment, governance, and strategic decision-making.

 

Institutions that actively embrace this shift—aligning technology, data, culture, and accountability—will not simply operate more efficiently. They will compete on an entirely different basis.

 

In the next phase of commercial banking, AI will not only accelerate credit decisions, but enable more intelligent pricing of risk, earlier detection of deterioration, and deeper, insight-driven engagement with clients.

 

Those that succeed will not just use AI to do the same things faster. They will use it to redefine what banking decisions look like altogether.

 

These themes — AI-driven credit decisions, data governance, organisational transformation, and the evolving role of risk professionals — will be at the heart of the nCino EMEA Summit, bringing together banking leaders and industry experts to explore what the next phase of commercial banking looks like in practice. Whether you are navigating the early stages of AI adoption or looking to accelerate an existing strategy, the Summit offers a unique opportunity to engage with peers, challenge assumptions, and gain actionable insight. Register today to secure your place at the conversation shaping the future of banking.

---

To learn more, please visit: info.ncino.com/EMEA-Summit-2026