In AI we trust?
In AI we trust? Only if it earns it
March 31, 2026
Content Director & Product Manager at Life Moments
Some have dubbed 2026 the year of “analogue.” Digital fatigue is real, social media has lost its lustre, and dumbphones are having a serious moment. In a world where going offline is the new status symbol, the title of a recent event felt particularly relevant: ‘In AI we trust?’ The question mark was certainly doing a lot of heavy lifting.
Despite the constant noise around agents, copilots, assistants and chatbots with increasingly friendly names, trust remains the big, unresolved question at the centre of AI. It’s not about whether the tech is impressive or if businesses are spending billions on it. The real hurdle is whether people actually trust it enough to use it when it really counts.
That was the recurring theme at the In AI we trust? event. In financial services, trust is not a “nice-to-have” – it’s the entire game. If you’re asking someone to trust an algorithm with their life savings, the bar is miles higher than asking Spotify to pick a workout playlist. The question that hung over the event was simple: if AI is the solution, what exactly is the problem?
Less hype, more reality
Jamie Bullock, Group Head – Client Experience at Fidelity International, provided a much-needed reality check. In a race where every company feels pressured to launch something AI-shaped, his reminder that “restricted scope is not a lack of ambition” was a breath of fresh air. Basically: do one thing properly – and build trust – before you start making big promises.
Naming a bot is easy – we’ve already got the likes of Freya, Cora, and Jude. But building something people actually rely on for real-world tasks? That’s the hard part. Dr Mark Thorpe, FMRS, Director at Truth Consulting, hit the nail on the head by asking firms to define their goal: Is it efficiency? Profit? Satisfaction? Or actually helping the customer? Too often, we start with the shiny new tech and try to work backwards to find a use for it.
Thorpe also described AI as “like the invention of the printing press on steroids”. A big claim, but it captures the scale of the shift. However, AI isn’t arriving into a world of open-armed consumers; it’s landing at a time when trust in institutions is already shaky. The Edelman Trust Barometer came up more than once, and for good reason. If people are already sceptical about the systems shaping their lives, they’re hardly going to hand over trust to AI without some convincing.
The human feedback loop
The discussion got most interesting when it moved away from the tech models and onto human behaviour. Prasad Chandrasheker, Global head of emerging technology strategy at Fidelity International, noted that “the concept of AI does not live in isolation.” The way AI interacts with people now matters more than ever, and trust is built through the experience over time. A tool can’t just work in theory – it has to feel clear, safe, and most importantly, offer a quick path back to a real human when needed.
Trust is built in those small, "it just works" moments.
Gabriela Rodriguez Elesgaray, AI Assurance Director at Deloitte, highlighted the growing demand for standards and certifications. But she was refreshingly honest about the downside: compliance and "responsible AI" frameworks are great until they become a bottleneck. It’s the classic corporate dilemma: everyone wants guardrails until they want to go fast.
Chandrasheker made a related point about accountability. Standards help, but when the technology is moving this quickly, business leaders also need to understand what responsibility really looks like in practice. And often, firms only discover how unprepared they are when they try to launch something – whether that’s because the data is patchy, the ownership is unclear, or the processes behind the experience just aren’t there yet.
In financial services, especially, a polished AI interface is one thing. Having the data, governance, and clear ownership behind it is another.
Is it Intelligence or “Impersonation”?
Paul Levy, writer and researcher at the University of Warwick, brought a healthy dose of scepticism, calling AI “artificial impersonation” rather than true intelligence. It’s a sharp take. His point was that customers aren’t exactly begging for more chatbots; they’re begging for help. There’s a massive difference between a slick digital interface and actually feeling supported.
While some suggest Gen Z and Gen Alpha would rather talk to a bot than pick up a phone, I’m not totally sold. My young daughter is constantly on the phone with friends and family. Younger generations might be tech-savvy, but they haven't lost the appetite for real human connection.
Practicality over magic
This is where Life Moments’ approach feels different. Our Design Director, Rob Winter, shared how we’ve been building our AI Coach around actual life journeys rather than just using it as a gimmick. In a day full of abstract questions, this practical angle mattered. Good AI should help people navigate the big, scary stuff in a way that feels supportive.
Oliver Lane, Founder and CEO of BehindLogin, added a final reality check: many chatbots are still glorified rules-based systems. A lot of what’s dressed up as “AI magic” is actually quite narrow. Consumers are smart – they can tell the difference between something that sounds clever and something that actually solves their problem.
The CMA’s recent report on agentic AI echoes this. The potential to save time and reduce friction is huge, but so are the risks of losing control. The bottom line? The law still applies, businesses are still responsible, and trust will only come through transparency.
Truth Consulting summed it up neatly with their “six faces of trust”: it has to be earned, transparency matters, people need to feel they matter, belief needs proof, value has to feel mutual, and the real test is how a firm responds when things go wrong.
So, in AI we trust?
Not by default. And certainly not yet. But maybe – just maybe – if it puts in the work to earn it.
What's coming up next?
AI was our starting point to kick off many more conversations we think are important to have. On AI, but also on trust relating to other topics prevalent in the Financial Industries in 2026.
Over the coming months, we’re continuing to explore what trust really looks like in practice. Next up, we’re turning our attention to Targeted Support, with both a webinar and an in-person session designed to unpack what it actually means to deliver it well.
At the same time, we’re keeping the AI conversation moving. From smaller roundtable discussions to more hands-on sessions, we continue to create space to interrogate what’s working, what isn’t, and what comes next for the industry.
And later in the year, we’ll be bringing the focus back to something ever so fundamental: the role of humans. Because for all the progress in technology, trust is still built between people.
Most of our events are invite-only, so get in touch with Daphne Toprek or Ben Leonard if you want to be in the know.
