The Rise of Agentic AI: Empowering Banks and Customers

Let's head back to Amsterdam, where key developments keep rolling in from day two of the Money 2020 Europe conference.

Not surprisingly, one of the biggest topics is how artificial intelligence is changing fintech at a rapid pace.

Patricia Wu joins us live from Amsterdam.

So Patricia, you hosted a panel today discussing agentic AI.

So tell us, what is it and why does it matter right now?

Hi there, Remy.

Well, I've got the perfect person to help us answer those questions.

It is Martin Muller, the head of AI for financial services for Europe, the Middle East, and Africa for Microsoft.

We did the panel together today about agentic AI and the future of banking.

So we're going to start with the basics.

What is an agent?

Very good question.

Look, the simplest way of thinking about it is when we had Gen AI, it was all about creating content, whether that's computer code, text, images, videos, you name it.

GI is about content.

Agents on the other hand side are about actually getting things done for me, doing things for me, getting it done for me, because they're actually able to use tools the same way that we are.

Think about them clicking, pointing, writing instead of me having to do it.

So that's really how you can best think about agents, independent or the technological argy bargy.

So what are banks doing with it?

I banks have already started doing Gen AI at the very first year this came out on enterprise grow, kind of 2023, so they've been really out of the gate faster among the first ones, and everybody is starting with three main ones.

So everybody's looking at contact center.

Nobody in the contact center likes to have to run around and find all the information themselves while they've got an inpatient customer waiting.

So AB and Amro, the big bank here in the Netherlands, they went out in June 2023 with a GAI solution that augments.

Center workers and many have followed since.

That's the first one.

Second one, everybody's looking in how can we augment the bank advisors.

Morgan Stanley in the US in May 2023 was among the first ones to do that.

Ever since then we had BlackRock, UBS, so many others do the same thing, effectively giving all of their bank advisors an army of specialists that they have in their pocket from the CIO to the product specialist to the research analyst, which reshapes how they can engage with customers.

And the third one is, how can we finally get the customer interface right?

Talking to a chap out of a bank is like pulling teeth.

If you don't use the right code word, it's not going to understand you, Amex instead of American Express, you're not going to get anywhere.

So Jenny, I changed that completely, naturally understands me.

Naturally understands the bank, but now that we've got agents, it can actually do things for you.

It can block your credit card, do the transfer for you, and that's what we see many, many people doing.

Wells Fargo in the US, Commerce Bank in Germany, for example, is even rolling this out with the GAI Avatar as kind of the new interface.

So those are the top 3 ones, and the leaders that have already done this over the last 2 years, they're looking to redesign the products.

Think about Agenta Commerce, what we announced with Visa, with Mastercard, with PayPal, and they're looking into how to gentrify the operations, especially on compliance with KYC.

So that's kind of the 3 + 2 patterns we're seeing.

So you know we keep saying this is the future.

How far are we talking with these new products and you know what is this long term vision look like and long term is it 3 to 5 years?

Is it next year?

I mean, things seem to be moving so quickly.

They are moving at an exponential rate and that's really amazing.

And it's not just that actually the technology is getting so much better.

There's another factor as well, the technology is getting so much cheaper.

You can basically draw a straight up exponential curve in terms of the capabilities of this technology, but a straight down line in terms of the cost per use of it, and that's really sparking a massive amount of adoption.

So we see financial services is actually leaning ahead of all other industries.

It's bigger when it comes to AI and a genetic adoption than healthcare, manufacturing, retail.

So that future is already here.

And when you're thinking about when do we really see those full scale agents that are really able to do entire jobs in the company.

And augment me so I don't have to do it.

I would say that's about 3 to 5 years maximum out with the first ones already live.

Even multi-agent systems already live today.

So we see the beginnings and over the next 3 years there's going to be a massive impact.

So we have just about 2 minutes left.

How is this massive transformation going to impact, I mean, value propositions, employees, customers take your pick?

I mean that's the very big question.

If you take a step back and zoom out.

In our private life we're all going to have AI companions that are going to be far more intimately connected to us than our phones ever were in digital transformation.

There's a Harvard analysis out just a few months ago that saw that nowadays the biggest use for Gen AIs is for people to give meaning, to be a friend, to even be a therapist.

So we're going to have all of these AI companions and suddenly they are so genetically empowered that they can do kind of anything for us.

So that means.

What is actually the value I'm still willing to pay companies for?

Banks are going to have to rethink that.

Banks will also have to think about what if everybody has an agent, and that agent talks to my bank agent.

I have to have parallel value propositions for an agent to agent economy, but still for a human to human economy and maybe even a third one, a human to agent economy.

So that's fundamentally changing the formula of what is value and how do we share value propositions.

And at the same time as you were hinting, if we have intelligence on tap.

Basically with agents in the enterprise augmenting us and extending us as employees, that's going to fundamentally redefine target operating models with agents driving dynamic workflow, so banks will have to think through that as well.

And I think those are the really interesting questions that we see the pioneers working through at the moment.

Right, so we have just one minute left.

If you had just one key takeaway for our audience, what would it be?

The one big one.

This is a race that the fastest learner is going to win.

Because you need to start with a start and you need to start doing it.

That's the only way how you learn standing on the sidelines while I'm doing it, because if you do it, if you build an organizational muscle on your tech side, they learn how to build this responsibly.

On the employee side, they learn how to use this technology to the max.

So this is the key thing start doing because this learning impact compounds, the ones that started 2 years ago already, they're that much faster, that much better at taking the next wave and the next one.

So that will be my big takeaway.

This is a race for the fastest learners.

Ah, I see.

OK, well, thank you so much, Martin.

That is certainly a good insight.

You heard it here.

That agentic AI train, as we use that analogy earlier, this transformation where a train ride, it has left the platform.

It is moving quickly.

Absolutely so.

Thank you.

Thank you.