How Can Banks Overcome Fear of AI Implementation Failure?

As banking continues to evolve around customers and their demands, AI has found a useful slot in serving both the bank and the customer. 

With the advancement of technology such as AI, banks must of course consider the risk factors when modernising their processes to include it. 

Is it the fear of failure, or the lack of confidence to push ahead that stops progress? 

In an exclusive Q&A with FinTech Magazine, Barb Morgan, Chief Product and Technology Officer for Temenos, explains how modernisation can benefit banks without making risky decisions. 

How is AI impacting banking – and what does the ‘AI bank of tomorrow’ actually look like? 

If you start with the customer, which is where we always begin, the shift is actually quite fundamental.  Today, banking is something customers do. They log in, navigate systems, fill in forms, wait for outcomes. It sits on their to-do list. 

In the AI bank of tomorrow, banking becomes something that works for them. The experience becomes inherently more natural and intuitive. Customers won’t be navigating menus or processes – they’ll be having conversations with their bank. 

They’ll ask a question in plain language and get an intelligent response that’s grounded in their context, their history and the regulatory environment around them. In that world, intent becomes the interface. 

But it also goes beyond conversation as AI starts to remove friction altogether. Agents can identify and resolve issues before the customer even knows there’s a problem – whether that’s a payment flagged incorrectly or a compliance requirement triggered by a cross-border transaction. 

So, the experience doesn’t just get simpler – in many cases, the friction disappears entirely. 

Moreover, where it becomes truly transformative is the move to agent-to-agent interaction. Customers will increasingly have their own AI advisor, a de facto personal agent acting on their behalf. 

That agent will interact directly with the bank’s agents to negotiate rates, verify transactions and manage financial outcomes, all within the intent and boundaries set by the customer. It’s still human-led, but with AI operating on their behalf. 

Importantly, this doesn’t replace what exists today. We’ve seen this pattern before. When web banking emerged, branches didn’t disappear. 

When mobile arrived, web didn’t go away. Different channels found their role depending on the moment and the need, and that continues to be true. 

So, the bank of tomorrow is inherently hybrid, meaning multiple channels, multiple modes of interaction, all working together seamlessly. Human, digital, conversational, agent-driven. 

That’s ultimately what customers will expect. And it’s what banks need to be preparing for now. 

If that’s what customers will expect, are banks actually ready for an AI-powered era?

Around 43% of banking platforms still run on technology designed before the internet existed, and as much as 80% of IT budgets can go to maintenance – not innovation or growth. 

That’s why shifting to modern banking products and platforms matters: it enables a step change in how banks run their businesses and deliver outcomes for customers.

But modernisation isn’t about replacing what works. It’s about changing how you change, so you can keep up, without getting left behind. The banks that have modernised are launching new products and experiences faster and responding to market shifts in near real time. 

Here’s the inflection point: transformational technologies like AI don’t just reward a modern platform – they increasingly require one.

If the core is a monolith locked behind batch processing, AI can’t access the right data at the right time. So, modernisation isn’t just about efficiency, it’s about creating the foundation that makes what’s coming next actually possible.

Banks are already applying AI, real-time analytics and automation to legacy systems – and getting some results. But there’s a difference between adding AI and truly supercharging it. To do that, your platform needs to be cloud-native with elastic scale. 

This means that AI can scale with demand – processing high volumes, running models in real time and scaling up (and down) as needed. On a legacy core, you’re constrained by fixed capacity. The technology may run but it runs with the handbrake on.

Platforms also need to be event-driven with real-time execution. When your platform communicates through APIs and real-time business events, AI can trigger action, like repairing a payment, flagging a transaction, updating a customer record, all in the moment something happens. 

On a legacy core locked behind batch processing, you’re working with yesterday’s data. You can observe but you can’t act in time.

Lastly, you need unified, accessible data. When transaction history, product configurations and regulatory frameworks are accessible across the business, not siloed by product line or geography, AI can reason with full context.

With disconnected data, decisions get made with partial information, and in banking, partial information is risk. That’s the modern platform and that’s what progressive modernisation builds towards.

If AI requires a modern platform, what practical paths can banks take to modernise?

In banking, modernisation can’t be a ‘big-bang’ event. You’re dealing with millions of live accounts, regulatory obligations across jurisdictions and customers who expect near-zero downtime. So ripping everything out and starting again isn’t just risky - it’s unnecessary.

That’s why we built our platform around composable banking. Each capability, deposits, lending, payments, is independent, deployed through APIs and event-driven connections. That means change in one area doesn’t ripple across the rest: you upgrade what you need and you keep what works.

From there, banks typically take one of two paths to a modern platform: Path one is to start with the core. Replace or upgrade core capabilities – composably and progressively – without disrupting what’s running. 

Path two is to start with the surround. Deploy point solutions in areas like wealth, payments or financial crime – capabilities that deliver value now, without touching the core.

Both paths are progressive and both lead to the same outcome: a modern, future-ready platform.

Banks are often described as having a ‘fear of failure’ when it comes to modernisation. Is that really what’s holding them back?

In banking, what gets labelled ‘fear of failure’ is often something more specific: fear of breaking trust. When you’re responsible for people’s money, operating under tight regulation and expected to be available all the time, the penalty for disruption feels immediate.

So the unlock isn’t ‘take bigger bets’ – it’s make change safer and more repeatable. The banks that move fastest are the ones that turn modernisation into a series of contained upgrades, with clear guardrails: what can change, how it’s tested, how it’s governed and how it’s audited.

That approach builds confidence – and momentum – because you can prove value quickly, learn and keep moving without putting the bank at risk. And over time, those incremental upgrades add up. 

The banks pulling ahead aren’t the ones taking reckless risks, they’re the ones that have built a disciplined way to modernise continuously, so progress doesn’t require disruption.

That’s the real cost of standing still – not failure.