Why banks are drowning in unachievable AI promises

The banking industry faces an AI problem. Vendors promise AI transformation but deliver only confusion. Generic chatbots require prompt-engineering. Multi-agent frameworks demand dozens of specialised tools. Both approaches create complexity that overwhelms rather than empowers.

 

According to Gartner, 54 per cent of AI projects fail to reach production, largely because of implementation complexity. That failure rate carries a particular sting in banking, with banks losing money and ceding ground to competitors who are investing in focused intelligence and building meaningful efficiency advantages.

 

The answer isn’t more AI tools; that’s just technology for technology’s sake.

 

The big complexity trap that many banks fall into

 

Walk into any major financial services technology conference and you’ll encounter two competing visions of the AI future. The first promises hundreds of specialised agents that, in theory, can be orchestrated into something meaningful. The second offers generic copilots that require huge internal resources – not great at a time when 75 per cent of banks are grappling with serious talent shortages, and the pressure to do more with less has never been higher.

 

Both visions fail because they treat banking innovation as something that can be retrofitted with general-purpose technology. This misses the point entirely when it comes to the difference between insight and action, and completely ignores the realities facing banking teams. 

 

For example, a credit analyst working on a covenant breach isn’t performing a generic information retrieval task.  A loan processor managing document exceptions isn’t simply moving files. And a relationship manager driving growth and retention is drawing on years of contextual knowledge that no amount of prompt engineering will replicate.

 

Quality over quantity is the answer

 

Banks need a different framework for AI adoption, one that begins not with the technology, but with the role. nCino calls this approach “intelligence without intimidation".

 

Our Agentic Banking Platform orchestrates role-based Digital Partners built around the actual job functions that drive banks – the executive, the analyst, the processor, the relationship manager and the account holder – with specific guardrails that keep every action auditable, every decision explainable and every workflow connected.

 

This goes way past generic copilots and instead understands the context in which a question is being asked, the relevant regulatory obligations to consider, the time pressures at play and the level of visibility required. This framework replaces complexity and confusion with empowerment and meaningful outcomes:

• Analyst Digital Partner handles covenant testing, financial spreading and early warning monitoring
• Service Digital Partner provides real-time client context at the moment of a conversation
• Processor Digital Partner tracks conditions, organises documents and validates compliance requirements
• Client Digital Partner enhances digital banking experiences directly through the banking interface
• Executive Digital Partner supports strategic decisions with portfolio intelligence and market sensing

 

When banks are competing on their ability to make faster, better-informed decisions, this kind of role-based advancement translates directly into competitive advantage. This isn’t theoretical. nCino’s Analyst Digital Partner is live, with early customers reducing commercial relationship review effort by 60 to 70 per cent, turning a process that once took days into one completed in hours.

 

Five agents, not 50

 

What’s revolutionary about this approach is its restraint. In a market where AI capability is too often measured by the number of tools on offer, the decision to build carefully designed agents rather than dozens of generic ones is a strategic and impactful choice delivering real differentiation.

 

The management overhead of large agent ecosystems is substantial and quickly creates conflicts and complexity, as well as limiting quality, coherence and adoption. nCino’s unified, role-based Digital Partners framework orchestrates workflows and tools behind the scenes, producing outcomes based on context and experiences by design.

 

It’s time to keep the promise

 

Customer expectations, shaped by digital-native experiences in every other domain of life, are now applied to banking relationships. The banks that close this gap first will not simply be more efficient; they’ll structurally shift the dynamic in terms of efficiency, competitiveness and experience.

 

This isn’t about replacing people with bots. Done the right way, the feedback loop between human judgement and machine intelligence grows tighter with every interaction.

 

Whether a bank is headquartered in London, Frankfurt, Amsterdam, Madrid or Riyadh, the opportunity to differentiate through operational intelligence is huge. The question is no longer whether AI can be useful. It’s who’ll embed it the right way to actually change how they operate, rather than simply adding to the technological burden and number of unachievable AI promises. 

 

Banking has always been an information business. What’s changing is the speed at which information can be synthesised, surfaced and acted on. With role-based Digital Partners, efficiency advantages build quarter-by-quarter, top talent mobilises and customers receive faster, more personalised experiences.

 

Is your bank ready to stop drowning in promises and start building advantages? The window is open. The technology is proven. The only question is whether your bank will lead the transformation or spend the next five years watching competitors do it first.

---

To explore how role-based Agentic AI is reshaping banking operations, click here