I loved the idea then and I still do.
But as AI has taken over the conversation, I’ve been questioning if ‘human touch’ is still the differentiator we once believed, or maybe hoped, it was.
To say that AI has complicated this idea is an understatement. ChatGPT, Gemini and Claude are now on speed dial, ready to deliver the elements of human touch better than many client service people.
That’s because AI is endlessly patient and attentive. It remembers what you tell it, doesn’t get irritated when you ask the same question five different ways and happily responds to your anxious text at two in the morning, with salient solutions delivered in the right tone of voice.
And, increasingly, AI can make you feel that it truly understands you and even empathises with you.
This doesn't mean AI feels empathy. It just means it has become remarkably good at performing many of the behaviours humans use to identify empathy.
As a brand leader and marketer, I find that fascinating, and a little uncomfortable.
Because brands, particularly in financial services, have spent decades telling people that our difference is personal service: ‘We know you’. ‘We understand you’. ‘We care’.
So, what happens when a machine can do a pretty convincing version of all three?
I don't believe the answer is to enter a pointless competition with AI to prove that humans are warmer.
AI is going to get better, and we should want it to.
Used well, it can make financial services faster, easier and more intelligent. It can analyse extraordinary quantities of information, identify patterns we might miss and remove friction from experiences that have frustrated clients for years.
The more interesting question is what becomes more valuable as those capabilities become commonplace.
I keep coming back to one word: accountability.
AI can give you an answer, make a recommendation and sound entirely confident about it - confidence is the one thing AI can produce in unlimited supply, even when it is wrong.
But it cannot stand behind that recommendation or own the consequences.
There is a well-established idea in AI ethics known as the ‘responsibility gap’: as machines play a greater role in decisions, it becomes increasingly important, and often increasingly difficult, to determine who is actually answerable for the outcome.
The machine itself cannot bear that responsibility. In effect, there is a huge (responsibility) gap between getting answers and giving advice.
We’re perfectly happy to use AI to help us know more, but when the consequences become significant, we still want a face and a name to take ownership of the transaction.
Perhaps that’s because advice from something with nothing at stake is ultimately still information, however beautifully it is packaged and delivered.
On the other hand, judgement, embedded in accountability, carries consequence. And consequence changes the nature of every relationship.
If accountability is part of the value that humans bring, it isn't simply enough to put a human face at the end of an automated process.
You must give that person the authority and agency to act.
A banker who knows your name, listens carefully and empathises with your situation but still has to say, “I'm sorry, that's what the system says,” isn't delivering human judgement.
A human who can only escalate isn’t accountable; they’re a face the client sees while the system decides.
I've spent much of my career thinking about what makes brands trustworthy, and one thing I have learnt is that what you say matters far less if the experience doesn't support it.
‘We care’ is easy to put in an advertisement. Giving people enough trust and authority to make a call on behalf of a client is considerably harder.
Most organisations love the idea of empowered employees but want the outcome without granting authority.
Real empowerment is more uncomfortable because it means giving people decision rights, accepting that judgement sometimes involves risk, and backing good people when they use it.
So how do organisations see people as more than profiles, segments and datasets?
Better data promises better understanding, better decisions and better experiences, but what if we backed people instead of just their data?
AI is forcing us all to grapple with what being human really means, what it demands of us, and why it matters.
Seeing somebody as more than data demands something more of you.
Someone to take ownership when it matters.
Someone to be deeply invested in the outcome.
For me, that’s where the conversation about humans and AI becomes interesting and urgent as accountability may become the most distinctly human advantage of all.
