Club @ Sibos
ARTIFICIAL INTELLIGENCE
at near-zero marginal cost. Institutions with
large and varied customer bases will likely
be the primary beneficiaries of an increased
capacity to tailor products at scale.”
A stumbling block, however, could be an
underestimation on the part of financial
institutions of the changes required to ready
their organisations for deployment of AI
technologies – an issue BNP’s Devambez cites
above. WEF found that institutions strug-
gle to make available the large quantities of
high-quality data required to successfully train
AI across their owned and unowned datasets.
In operations, many valuable applications of
AI require complex, deep and broad-reaching
integration into the business, not just simple
“bolt-on” implementations. Moreover, AI
fundamentally redefines the role of talent
in financial institutions, and often requires
human capital to change at a pace that exceeds
any past transformation. Finally, current
regulatory frameworks were built based on an
increasingly outdated model of the financial
ecosystem, creating significant uncertainties
for institutions seeking to employ cutting-edge
implementations of AI.
These two final points are also cited by Andy
Burner chief information officer at software
and managed services company SmartStream.
A stumbling block for AI could be the regula-
tory environment, he warns. Human verifica-
tion is still a requirement for most regulators
and the fact that an AI- or ML-based applica-
tion can make a decision, without saying how
it came to that decision, could be deemed a
risk. Like Devambez, Burner advises financial
institutions to try to understand AI. “It can be
very difficult to identify use cases and in many
ways it is not just about a use case, it is about
understanding where to use a new tool. It is
very difficult to find data scientists who also
understand the financial services business.”
Working with financial technology start-ups
isn’t always the answer either; often such com-
panies don’t have the same agenda as the bank
with which it is working.
However, he points out that banks, finan-
cial institutions and vendors have optimised
workflows and products a great deal during
the past few years; it is now difficult to further
optimise conventional technologies. “New
technologies such as AI, machine learning and
blockchain allow for further cost-effective
optimisation,” he says. “We hear a lot about
these technologies and cost reduction, but it’s
also about the superior insights into data that
can be created that will greatly improve the
user experience.” In cash and liquidity man-
agement, for example, AI can provide insights
into years of data, improving forecasting and
enabling better leverage of cash and liquidity.
Industry analysts Celent believe that AI and
robotic process automation (RPA) tools are
finally making significant inroads in post-trade
White paper
The transformation of
the European payments
landscape
To read the full report go
to swift.com/white-papers
operations, automating many manual process-
es. In its report, The Next Generation of Post-
Trade Technology, Evolution to Revolution,
Celent identifies several use cases for AI in
risk and compliance, surveillance, reconcilia-
tion, and order management.
“The development of commercial use cases
of AI and RPA for capital markets started three
years ago, and there are many front office
commercially available use cases now in areas
such as KYC, AML, trade surveillance, and
natural language generation of standard in-
vestment research reports,” say report authors
Joséphine de Chazournes and Arin Ray.
RPA systems have been used initially to auto-
mate highly manual and repetitive processes to
reduce costs and as an alternative to outsourc-
ing. Celent states that this could lead to some
headcount reduction, but the more important
objective is to enable current resources to
process enormous amounts of data efficiently
and accurately.
The loss of jobs to automation is a concern
across many industries – and has been since
the industrial revolution. The WEF report
states: “As the global financial system under-
goes transformation, institutions, regulators
and policy officials must be proactive to man-
age the large-scale displacement of labour, as
well as develop modern tools to manage new
ethical uncertainties created by automated
decision-making.”