AI in Banking: Why Verified Business Intelligence Is the Foundation for Scale

AI is changing how banks manage risk, serve customers, and pursue growth.

Sara de la Torre, Vice President of Global Financial Services and Alliances at Dun & Bradstreet, believes “global banking is going through one of the biggest changes we’ve seen in history.”

This is, by and large, due to the convergence of AI, cloud, and data, coming together in ways never seen before.

But as AI capabilities become more widely available, adoption alone will not create lasting advantage. The difference will lie in how banks apply AI to business priorities such as increasing financial returns and mitigating risk, how they support it with verified business intelligence, and how they scale successful applications while maintaining their reputation. They must also benchmark their progress against competitors throughout.

What AI Maturity in Banking Looks Like

Evident is an intelligence and benchmarking platform whose AI Index assesses how effectively leading banks develop, deploy, and govern AI. It offers a view of who is leading the race and also helps reveal how the sector is evolving.

When Evident started the index three years ago, only two banks actively communicated their AI spending, deployments, and returns. Now, more than 17 banks share that information. So where are they focusing?

The latest Index indicates that leading banks made AI an early strategic priority, proving value with pilots in selected areas before expanding.

“They started with a succinct strategy, focusing on areas where they generated success, then broadened out,” explains Alexandra Mousavizadeh, Co-Founder and Co-CEO of Evident.

According to Mousavizadeh, around 85% of AI use cases are still internal. Internal tools and workflow automation have often progressed first because they can be more closely controlled. Processes including Know Your Customer (KYC), Know Your Business (KYB), sanctions screening, and fraud detection have increasingly become part of banks’ core AI efforts.

Going forward, the greater opportunity for differentiation lies in customer-facing applications that help banks identify growth areas, expand relationships, and deploy capital more effectively.

How Banks Turn AI Efficiency into Competitive Advantage

Mousavizadeh notes that leading banks are deploying AI across 18-20 lines of business including table-stakes, and emerging and frontier use cases. Efficiency has greater strategic value when it supports revenue growth as well as reduces costs.

A key example of where agentic AI is delivering returns already is within onboarding processes – across both compliance and financial risk assessments. What would previously take months is now being done in days, or in some low-risk cases, minutes. Bringing customers into the bank sooner may allow them to begin using products and generating revenue earlier.

For example, using data from the D&B Commercial Graph^™, a KYB workflow built within Anthropic’s Claude^® can reduce a simple, low-risk onboarding case from days to minutes.^* Similarly, using Dun & Bradstreet’s verified business intelligence within Databricks can increase bad-capture rates in a commercial loan portfolio from 30% to 38% and contribute to a $6 million reduction in bad-debt write-offs.^*

However, the opportunity extends beyond individual processes and internal workflows. “Where differentiation is happening, and where the competitive edge lies, is at the front line,” says Mousavizadeh.

AI-supported tools can help relationship managers understand corporate connections, identify unmet needs, and recognize opportunities to increase wallet share. This could be through customer self-service or by serving sales and customer service teams with faster and more accurate business intelligence.

It could also support targeting and prospecting for wealth management or help banks to model optimizations to credit policies, collections, and bad debt positioning – uncovering opportunities for growth and more effective capital provisioning.

Banks that cannot identify opportunities, onboard customers, or deploy capital as quickly as their peers may miss valuable routes to commercial growth.

Why Verified Business Intelligence Is Critical for the Future of AI in Banking

To make these pilots and production workflows a success, banks are making broad and deep investments in hyperscaler partnerships, platform architecture, and talent.

However, every AI-supported banking decision depends on data. Our latest AI Momentum Survey of 10,000 businesses found that over three-quarters of enterprises report some measurable ROI from AI, but only 6% say their enterprise data is "fully ready" to support AI at scale.

“The challenge now is that AI adoption has outpaced data readiness. That is why only a few organizations have turned pilots into P&L-level ROI," says Gary Kotovets, Chief Data and Analytics Officer at Dun & Bradstreet. "Today's frontier models are extremely capable, but getting the context right is the key to effectiveness. Grounding AI in verified information, so facts can be confirmed and integrations can be established, is fundamental to adoption of agentic workflows."

For example, a business entity that a bank is looking to target, onboard, or loan to may be represented differently across onboarding platforms, loan origination systems, CRM tools, and regional databases. Parent companies, subsidiaries, and owners can also be difficult to distinguish.

These inconsistencies can affect identity verification, sanctions screening, lending limits, and portfolio management. AI can process information quickly, but it cannot compensate for every gap, duplication, or mismatch in the underlying data.

“Banks aren’t really worried about the models. They’re worried about the trust that sits within them,” explains Jennifer Moore, Vice President and Global Industry Practice Lead for Banking at Dun & Bradstreet.

Banks also need confidence that AI-supported decisions can be explained, defended, and aligned with regulatory expectations, helping protect customer trust and institutional reputation.

Dun & Bradstreet’s D&B Commercial Graph^™ comprises 650+ million business records (more than any other provider) across more than 250 markets worldwide. Anchored by the Dun & Bradstreet D‑U‑N‑S^® Number (a unique and persistent identifier), the D&B Commercial Graph^™ provides a verified commercial identity foundation that can help enterprises deploy AI at scale.

It structures and connects business identity consistently across systems, departments, and geographies. This provides a data context layer that can help banks identify prospective account holders, support more precise lending-risk decisions, and manage KYC risk alongside disciplined revenue growth.

How Digital Governance Enables AI at Scale

To capitalize on AI, banks need to give functions and regional teams room to develop relevant applications. But decentralized experimentation can create fragmented data, duplicated investment, and inconsistent controls. Agents are often deployed one-by-one, without interconnectivity.

A federated model is recommended to drive governance that keeps shared standards for data governance, security, explainability, and model risk, while allowing local innovation. This helps banks to scale AI safely.

When it comes to data governance, a single trusted data context layer such as the D‑U‑N‑S^® Number and D&B Commercial Graph^™ at the infrastructure level is key. Think of it like a golden thread used across legacy tech, ERP/CRM and other workflow systems, data lakehouses, hyperscalers, agents, and more. This supports a model-agnostic approach, giving banks greater flexibility as AI providers and capabilities evolve.

From Experimentation to Business Value

AI maturity should be measured by a bank’s ability to apply AI to material priorities and move successful applications into production responsibly.

For senior banking leaders, the value lies in whether AI can improve productivity, reduce costs, support revenue growth, and increase market share without compromising regulatory confidence or reputation.

The banks most likely to create lasting advantage will be those that connect AI to verified business intelligence, support explainable decisions, and scale successful applications across the organization.

Find out how to ground AI models, agents, and workflows with Dun & Bradstreet's verified business intelligence – anchored in the D‑U‑N‑S Number.