ENGLISH

09-SAS V3.vtt

We're here at Money 20/20
USA for FinTech Futures.

I am pleased to say that today we are
here with friends from SAS and I would

like to talk a lot about generative AI
and take a hard look into the future.

Looking Down the road at 2035,
things to come, and what are the

Diana, would you like
to introduce yourself?

I'm the Global Lead for Fraud,
Risk, and Compliance at SAS.

Basically my background is to
take a look at our solutions and

figure out how they fit in our
different geos across the world.

And how we can position them better
for our clients and future clients.

So I'm Theo, founder of
Unconventional Ventures.

I also hold a podcast called
One Vision and tell stories of

Alright, so we can't have a
conversation in this room without

It's become the impossible
elephant in the room.

Let's start off with talking about
where we're seeing use cases right now.

What have you seen, even on a global
scale, about where it's actually

So I think starting to, it's really
interesting I've had a lot of

conversations about how it's being used
by the bad actors, because, again it's

not that I don't think our financial
institutions want to not embrace it, but

I think it's hard to embrace it if you
don't understand how it's being used by

those bad actors it's everything you have
things from, I will, I've been saying

to people, chat GTP has made a fraudster
out of even the most common individual.

Really, just being able to access
something which can be used for good and

for bad, but being able to do things like
deep fakes, the amount of voice analysis,

and the way that people are mimicking
that now when they're doing call centers.

Or anything when you need a voice and
authentication, but really just even

down to how they can get in through our
financial institutions and compromise

And the problem with that is a lot
of the financial institutions haven't

really taken a look at their models
and rules to see how they can combat

all these bad actors with this new
technology that's coming out right now.

I think that's the one aspect of Chat
GPT that I'm so scared of, because as

Diana, you said, you can clone someone's
voice in three seconds or less, and we

already start seeing use cases of where
people get scammed because they thought

We really need to come up with
a way to figure that out really

We are seeing financial institutions
starting to use that to help their

CSR agents in the call center,
being able to get to information

faster and to be able to help the
callers in, a more efficient manner.

And recently we start hearing
wealth advisors using that as an

internal tool for wealth managers
to look through information that

So those are good developments
that we start seeing banks

It's experimenting with that, but one word
I always use is just be careful, right?

Innovate, but innovate
responsibly and carefully.

Part of that responsibleness has
to fall into how we're interacting

So what have you seen as financial
institutions almost are at a point where

they have to adopt faster technology to
combat what's happening on the fraudsters?

How are they staying in
compliance and what's happening?

I think what I'm seeing a lot of and what
I do really like to see from financial

And bringing together of so
called committees to work

Financial institutions and you hear
that, regulators, compliance, you

just take a step back and you freeze a
little bit, but I think now, this open

partnership that has happened, and it's
across the globe, it's really just not

in one specific country, I think we're
seeing it more in the EMEA markets.

Then anything else, but this partnership
where they can learn together, but also

work with the regulators as we're trying
to implement AI, as we're trying to

work through things so that we can work
in lockstep together versus let's just

say we're three years down the road,
somebody implements a system, they try

to go for AI and the regulators like,
hey, you shouldn't have done that.

That's wrong because you've crossed the
line now on all these different levels.

One thing I do is seeing CFP, for
example, coming out and say, Hey, wait

You need to make sure that
you can explain your model.

You can explain whether or not you're
accepting or rejecting someone and why.

And, it goes back to when we think
about using AI in financial services,

for example, in lending, a lot of
the data that we use to train the

How is that being manifested in the
training data that FISF is using to

determine what you can and cannot do,
and going back and saying, you know

what, this is the transparent model
we're doing, this is how we explain

I think that is incredibly important
and I hope we can do more of that.

It's a big picture scope that you're
bringing up, Theo, because you need,

I think, to have humans involved in
the AI process all the way through.

So how does a banker know when to start
being a personal touch and when to

I think what we're seeing is a lot of
the collaboration again, and what I

SAS at Money20/20 USA 2023 – The growth of generative AI in financial services