These are not illogical decisions. 

But despite the might of machines, only humans can underpin critical trust-defining moments: counseling complex decisions and life‑changing financial choices. More than ever, these ‘moments of truth’ define customer relationships. Whether simple ‘if-then’ rule-based or AI-powered, the wide-scale automation of first‑level service is often where customers first lose faith - pressing 1, then 2, meeting an uncomprehending bot and falling out of the system.  

During the early implementation of “Zero Hands”, too many institutions overlooked customer–employee interaction in mapping and modifying major processes. As the consequences become clear, it is time to step back and redesign how trust is delivered at scale. To move beyond tactics to holistic thinking. 

Lighting the customer path

As the digital wave sweeps across financial institutions, executive decisions increasingly hinge on understanding what not to automate. Identifying where human intervention matters and where it doesn’t. This demands a deep understanding of where those moments of customer trust sit.  

But a customer journey can contain thousands of interactions. It can be difficult to know where to start (or stop). Historically, process and value‑stream optimization happened at lower organizational levels. But the complexity, strategic stakes and risks all mean that top executives must now have their finger on the pulse, rather than simply pushing AI and technology, or trying to “turn us into an AI bank.”  

Without human fallback, institutions run a spiraling risk of onboarding the wrong customers, over- or underselling, discrimination or bias. And, if ‘Know Your Customer’ (KYC) tools are generally robust, full automation is premature. Fraudsters are becoming more creative - using AI.  

Executives, not just technicians, must take the helm together

My advice to boards and CEOs is simple: be explicit about the moments where human judgment is non‑negotiable. Use AI to strengthen, not replace, them. 

AI integration demands clear oversight. A strategic view from the CEO of what the organization will touch, versus the ‘off-limits’ moments of truth. A Chief Transformation Officer working closely with executives to ensure coherent deployment. Divisional directors with a firm grasp of their customer processes. Questions should be asked: which departments are no longer relevant? Do we need new ones?  

Here is the problem.  

Traditional organizational models in financial services are silo‑operated: CTO/CIO for technology, COO for operations, CHRO for people, and so on. This is no longer viable. Dilemmas could be solved by the rich data pools to which financial institutions should have access. But too many are still running old operating models, and customer data is departmentally fragmented or poor.  

There is no overarching system to optimize the full customer journey. So, customers are passed around, confused, frustrated, even misinformed. To optimize automated credit decisions, for example, demands collaboration between IT, customer, and process teams.  

This isn’t a project; it’s a multi‑year journey. No one can foresee what tools will exist in 18 months. The speed of development - think of how quickly Claude emerged - renders traditional project‑based approaches obsolete. Layering AI tools and advanced IT systems atop ageing operating models is like expecting optimal performance from an old engine with new parts. Consider anti‑money‑laundering processes: instead of redesigning the operating model (who does what, why, and when), AI tools are simply grafted onto outmoded structures.  

AI tools are evolving faster than regulation

Governance is a central pillar of AI: the rules, processes and frameworks underpinning how it is developed and deployed (ethically, safely, complying with legal and societal standards). Financial services are acutely regulation-aware. But in 2025, only 51% had assigned a C-suite manager to AI governance, say the IIF and EY, with only 45% facilitating cross-functional management of AI risks.  

Leaders must navigate a fragmented global picture – one that is shifting from voluntary principles to enforceable rules, with diverging requirements on risk, transparency, data governance, and accountability. Whilst the EU, US, and China are leading, over 70 countries have now installed AI policies.  

The threat doesn’t lie in AI adoption, but in removing humans from critical trust points.  

The opposite is also true. 

Trust moments are the lifeblood of a financial services firm. As The Silicon Review puts it: those that view AI’s 11.7% contribution in terms of pure cost savings face a social and operational backlash. 

Fortunately, this is a two-sided coin.  

Traditional institutions are full of talented people. Automation is an unprecedented opportunity to elevate the value proposition. One that blends the very best of people and AI. Competing, too, with nimble, AI‑native entrants.  

Trust is a core strength. Failing to capitalize on it means falling behind. Nourishing it will set your organization apart - sustainably. 

----------

Sources

[i] The Silicon Review (2025). ‘MIT Study: AI Can Replace 11.7% of US Jobs.’ (01 December, 2025).

[ii] Ernst 2024 EY European Financial Services AI Survey. 116  financial services executives from 106 private and publicly listed companies.

[iii] IIF-EY Annual Survey Report on AI Use in Financial Services 2025. Survey of 61 financial services institutions across 8 regions. 7th in a series beginning 2018.

[iv] Hoskins, P., (2026). ‘Standard Chartered to cut thousands of roles as AI use increases.’ BBC, (19 May, 2026).