It feels like AI has taken over in the last few years. It’s being used for everything from researching and proofreading to generating content and bouncing ideas off, even as a therapist and companion. Somewhat alarmingly, I’ve seen ChatGPT described on social media as “the only man who won’t cheat on you”, which is rather telling about our changing relationship with technology.
Closer to home, GenAI is being adopted across the workplace, and it’s featuring in client briefs with increasing frequency. And not just in terms of the research methodology – Pegasus has more than one project ongoing right now testing possible new financial services, all based on AI. It’s everywhere, and it’s happening fast.
Truthfully, as much as I enjoy playing with GenAI tools, I can’t imagine anything worse than a chatbot trying to ‘help’ me with my banking. After all, many companies are still trying to get basic customer service chatbots right. We’ve all experienced the horror of being trapped in an endless loop of “I’m sorry, I didn’t understand that” when trying to renew an insurance policy or ask a basic question about your account. Based on my experience talking to my own bank’s chatbot, the improvements in publicly available GenAI tools have not significantly increased their capabilities. If you haven’t resorted to repeatedly typing “talk to a human” until you get through to someone with a pulse, have you really lived?
But that’s just me – what do other people think? Turns out, quite a lot, actually.
Public perceptions of GenAI
A 2023 FIS survey1 revealed that nearly 30% of UK consumers do not trust generative AI at all. Additionally, 74% are unaware if their bank uses GenAI, and opinions are split evenly on comfort levels regarding its use in banking operations. So, there’s clearly a mixed picture out there, with most people simply unaware of what’s already happening with their own banks.
Okay, that was in 2023, and things have moved on a little since then. But a more recent study2 revealed that less than half (42 per cent) of the UK population are willing to trust AI, so there’s perhaps some way to go in terms of public confidence.
What did our own research reveal?
I wasn’t entirely convinced, so I decided to ask people myself via Potentia Insight’s Quick Pulse service. What did we find?
First, the basics: 46% of UK consumers use GenAI tools at least weekly, increasing to 79% of 18-34 year olds. For those aged 55+, usage rates drop to just 17%, with a third of this group saying they never use GenAI. So far, no surprises.
What about financial services, then? This is where our findings get more interesting.
Two in three are comfortable with their bank using GenAI services (such as chatbots and automated advice) to support their account and services, although the 55+ age group are notably less so. That’s quite a shift from the 2023 FIS study, where only half were comfortable. Although given the rise in AI’s stock and prevalence over recent months, this isn’t unexpected.
But this is where the data really surprised me. 82% of people we spoke to said they would trust financial advice provided by a GenAI tool to some extent, and 29% said they’d trust it as much as advice from a human. That’s almost 3 in 10 people, and rises to half (51%) of 18-34 year olds! I’ll be honest, that made me do a double-take.
Looking at the data, I’m forced to admit it – I’m falling behind the times. It’s a somewhat terrifying concept, but it did get me thinking: we’ve seen this before, haven’t we? When digital banking first emerged, younger consumers were quick to adopt while older ones hesitated. Eventually, most caught up, but only after banks worked hard to build trust, accessibility, and clarity around what they were actually doing with people’s money.
Can financial services get AI right?
Now we might be looking at the next big shift. Some of our clients are already starting to adapt, as is clear from the research briefs coming through. But it’s a delicate balance. To keep up with younger consumers, financial institutions will need to embrace GenAI tools – not as gimmicks or flashy add-ons, but as trusted, well-designed features that solve real problems. At the same time, they must not leave behind older customers who still want a human face, a clear explanation, and the reassurance that their money isn’t being handled by some inscrutable algorithm.
So, the question isn’t whether financial institutions will use AI, or even whether consumers will accept it – the answer to that is increasingly clear. The real question is: can financial services design AI experiences people actually want to use? And just as importantly, can they do it without completely alienating the customers who still prefer to speak to an actual human being when their card gets declined at the supermarket checkout?
If you’re looking to design products or services using AI, whether you’re convinced it’s the future or still a bit sceptical, we can help you test what works with real people. Get in touch if you’d like support navigating this brave new world – we promise our research methods are still reassuringly human.
2https://kpmg.com/uk/en/media/press-releases/2025/04/majority-of-uk-public.html
