The promise of artificial intelligence (AI) in financial services is not only to smooth customer sign-ups, but also to reduce fraud and money laundering. Progress has been slow, however.
“No-one is doing anything absolutely groundbreaking with AI. It is embedded in a ton of stuff, but we’re not seeing the mainstream banks in Ireland rolling-up their sleeves and investing in it,” said Mark Kenny, client director of retail banking, Expleo Group.
That does not mean that AI is not being used, however. Across Europe, the insurance sector has started to use it to help eliminate fraud.
“In the insurance industry we are seeing a lot of it in fighting fraud, and there is also the use of technologies such as telematics, the connected car, internet of things in home insurance,” said Kenny.
Nonetheless, even insurance has some risk aversion when it comes to deploying novel technologies. For example, Life/400 software is still running on AS/400 servers, as it was two decades ago. This is thanks in no small part to excellent forward-compatibility, but it does pose the question: is an opportunity for renewal being missed?
“The argument is there that core banking systems can and should be replaced,” said Kenny
The march of technology is relentless, of course, but is the wider financial services on the bleeding- or the trailing-edge?
In fact, neither position would be ideal for financial services. Too much is at stake. Nonetheless, there is certainly an opportunity to get ahead of technology-driven social change. In doing this, financial services can position itself as a leader rather than a follower.
“In automotive [insurance] it’s more a case that the telematics is driving the industry rather than the other way around,” he said.
Globally, Expleo counts among its customers major enterprises including Airbus, Barclays, BP, Commerzbank, Credit Suisse, Daimler, Deutsche Bank, Deutsche Post AG, Volkswagen AG, Eurobet and Zurich Group.
From this perch it has seen that new tech like AI and machine learning (ML) is being deployed, even if sometimes this work is going on more at the fringes.
For example, applied to optical character recognition (OCR), it can automate significant aspects of the customers’ application processes.
“We do see AI and ML embedded in banking and in the OCR world for customer on-boarding, speeding up applications,” said Kenny.
“We do also see it embedded in core systems so help create more seamless customer journeys, with some element of self-learning to keep that process moving.”
One of Expleo’s core functions in financial services, application testing, also benefits from an AI boost. The result is a deployment speed boost that helps banks move forward in their digital transformation strategies.
“Traditional testing can create bottlenecks, but in the agile world people want to get their things out there faster. ‘Deep learning’ can help us get them out there,” he said.
“We’re also working with banks to leverage AI to help with day-to-day operations and around incident prediction, collecting data from a variety of sources to help with the streamlining of operational environments.”
Structured and unstructured data is also widely analysed for anti-money laundering and fraud prevention purposes.
“AI can provide the data or machine learning to find patterns for good and bad applications, and the systems can then continue to learn. There will still have to be human insight, and the decision-making process and reason has to be explainable.”
The dark side is also an issue, though: hackers are already beginning to deploy AI tools.
“Hackers have access to AI tools and are leveraging them to gain access to enterprise systems,” said Kenny. “The flip side of that is using AI to stop it. AI needs to be built into tools rather than sitting outside of it or on top of it.”