Banking executives are rapidly rolling out AI initiatives, but a growing gap between strategy and execution is now emerging as the real challenge, according to nCino CEO Sean Desmond.
Desmond describes a sense of “speculation fatigue” building across the industry, where AI dominates conference agendas, keynote speeches and corporate communications, yet too often fails to translate into meaningful operational change inside financial institutions.
While adoption continues to accelerate, he argues that the real constraint is no longer awareness or intent, but delivery.
The issue, he suggests, is not whether AI will transform banking, but how leaders are actively embedding it into their own decision-making and operating rhythms. That question became personal earlier this year when Desmond turned the lens inward and asked what he himself was doing to modernise the way nCino operates.
From boardroom question to working prototype
The turning point came during preparations for a board meeting on March 16, when Desmond reframed AI not as an abstract enterprise initiative, but as a leadership responsibility. Rather than focusing solely on enterprise-wide deployment, he asked a more direct question: what AI agents were actually supporting him and his executive team day to day?
That line of thinking became the brief for an internal experiment — one grounded in real strategic context, including product vision, platform strategy and competitive positioning.
Desmond is clear that this was not an exercise in so-called “vibe coding”, where users prompt AI systems loosely and accept outputs with limited scrutiny. Instead, he emphasises a structured approach in which human leadership remains firmly in control, while AI acts as a high-speed analytical and synthesis layer.
When the system suggested delegating certain responsibilities to a chief of staff function, for example, he rejected the idea immediately, reinforcing that strategic judgment remained entirely human-led.
The outcome was a tightly scoped executive AI workflow designed to accelerate thinking rather than replace it.
A CEO agent stack built for execution speed
What emerged from the exercise was a CEO agent stack designed to support executive decision-making across board preparation, strategic planning and operational oversight.
Desmond describes it as a “thinking partner” that compresses execution cycles, maintains consistency across documents and surfaces insights that would otherwise require multiple layers of internal coordination. It does not generate strategy autonomously, but instead helps structure and accelerate work already defined by leadership.
In practical terms, tasks that would previously have taken several days of preparation — including board materials, executive notes and technical build documentation — were completed in around 90 minutes.
The key distinction, he argues, is that judgment never leaves the CEO. AI increases velocity, but leadership remains the controlling function.
Within 30 days, the system moved from initial brief to live deployment across the executive leadership team.
From tooling to operating model
For Desmond, the broader significance lies not in the tool itself, but in the operating model it represents. At nCino, this is supported by what the company calls its Agentic Operating System (AOS), which provides the underlying framework for coordinating AI agents across financial workflows.
Each morning, the system now delivers a curated executive briefing, combining external market intelligence with internal operational signals. Rather than overwhelming leadership with data, it filters information down to a small number of priority actions designed to guide the day’s decisions.
Desmond argues that this shift fundamentally changes executive attention — reducing noise, increasing focus, and tightening the loop between insight and action.
Leadership accountability in the dual workforce era
A central theme in his argument is leadership alignment. As banks deploy AI tools across lending, credit analysis and operations, Desmond questions whether executives are matching that adoption at leadership level.
He argues that credibility becomes difficult if senior leaders expect employees to work alongside AI systems while not using equivalent tools themselves. This, he suggests, is becoming a defining issue in how digital transformation programmes are perceived internally.
He also points to the emergence of what he calls a “dual workforce”, where human employees and AI digital partners operate side by side as part of standard banking operations.
Motion versus momentum in banking AI adoption
Desmond draws a clear distinction between activity and impact. Many institutions, he argues, are still operating in “motion” — running pilots, launching initiatives and holding governance meetings that create the appearance of progress without generating compounding value.
“Momentum”, by contrast, is when each AI-enabled workflow improves the next, building institutional intelligence over time and creating measurable performance gains.
He references nCino’s AI in Banking Benchmark, presented at nSight 2026, which found that 84% of banking executives are deploying AI at an enterprise level. However, a significant proportion remain focused on adoption rather than return on investment, suggesting that execution maturity still lags behind ambition.
Closing the execution gap
The core message from nCino is that AI capability is no longer the limiting factor — execution is.
Desmond argues that the organisations pulling ahead are those that have moved beyond experimentation into embedded operating models, where AI is integrated into daily workflows and continuously improves through use.
He adds that nCino’s Agentic Operating System and executive digital partner framework are already available to financial institutions seeking to make that shift.
The final point is simple but direct: the technology is already in place. The question now is whether banking leaders are willing to use it themselves, not just deploy it to others.
Read the full blog from nCino here.
Copyright © 2026 FinTech Global
