The Intermediary – August 2025 - Flipbook - Page 28
T H E I N T E RV I E W
SBS
a hard task for many, not least when this might
be seen as a hygiene factor – shifting in order
to continue functioning smoothly, rather than
to create any clear business or profit gain.
To make this transition easier, Bierry says:
“Those legacies can stay, but they can stay
with conditions – adopting a standard, opening
up those legacy systems and adopting all the
necessary APIs, and bringing down the time to
market expected by financial institutions.”
He adds: “We are already seeing the benefit
for 18 or 19 of the building societies we
work with today, which have adopted these
standardised systems. Then, they can focus on
the product catalogue and how they engage
with clients.”
Standardisation is also a benefit for those
smaller businesses that may not have the
resources adopt the most advanced, bespoke
technology. Instead, without having to invest in
tailored specifics, smaller businesses are better
able to compete and focus on growth, which
has a broader positive effect on the health of
the overall market, through better competition.
Bierry says: “This is not an easy decision
to make at first, but once made it is an easy
execution. A building society with less than
£1bn of assets under management cannot
afford to do a two-year IT programme, so
their only way to survive is to stay focused on
growth and business development, how they
want to be seen as different as a niche player,
rather than spending money and time on IT and
back office processes.”
AI insecurity
Any discussion of the tech underpinning the
modern lending landscape inevitably rolls
around to artificial intelligence (AI), either with
excitement for its capabilities, or trepidation
about the brave new world it heralds.
For Bierry, the discussion is moot until
financial services gets better at leveraging data.
He says: “Historically, financial institutions
don’t like to use the data they already have. It
comes from their DNA – a history where people
were coming to their bank in person to deposit
physical money is still part of their processes.
“So, the first step is accepting the use of data
– with or without the use of AI.”
When it comes to the dawning use of AI
in financial services, whether a reality or a a
future discussion, the biggest fear tends to
be around the safety of customer data. This is
not unfounded, and the growing conversation
around public large language models (LLMs)
is fuelling this concern. However, not all
systems are created equal, and SBS has its own
26
The Intermediary | August 2025
proprietary AI system, allowing it to be used
locally by a lender, without either pulling in
unvetted external data or allowing internal data
to be dispersed elsewhere or “exposed to the
public cloud.”
While this format is more limited in scope
than LLMs such as ChatGPT, it is about finding
systems and tools that are fit for purpose
and work with the “real use cases lenders are
expecting to deliver to the market.”
With the realities of audit trails, GDPR
compliance and Consumer Duty requirements
pressing heavily on all financial services
businesses, those in this market that do not
enter into the tech and AI conversation are not
just going to struggle to compete, but may face
nasty compliance shocks along the way.
When it comes to this progress, Bierry
adds: “For those of our clients that have not
yet matured into using these systems, it’s not
because of privacy concerns, it’s because they
need to have the right data. They need to decide
the way they want to leverage that data, and
they need to decide what kind of use case they
would like to go to the market.”
Naysayers in all markets will warn against the
use of AI, of course. Bierry notes that often this
is about a misunderstanding of the solutions
out there, and a lack of awareness that closed
systems do exist.
On the subject of growing this awareness, he
says: “This is going to take years. And the speed
of the market is not going to allow for the fact
that teaching people is going to take years.”
Part of the solution to this is transparency
and communication, not least because AI comes
in various diverse forms, many of which the
average person interacts with throughout their
day without even realising.
For example, it is important to distinguish
between generative and predictive AI.
Predictive AI is widely used – from making
shopping recommendations to risk modelling
and identifying suspicious patterns in fraud
detection. Generative AI, meanwhile, is the
“next stage” in the evolution. This is centred
around the actual creation of content, from
personalised marketing through, potentially, to
mortgage agreements and KFI documents.
The vast majority of the use cases in financial
services at this stage are predictive. Bierry
points to the use, for example, in preventing or
anticipating defaults.
He says: “It’s a lot better to engage with your
client three months before a default, proposing
something to help them, rather than waiting
for the default and then entering into a very
difficult conversation.”
