Artificial intelligence (AI) is no longer just a daydream reserved for the imaginations of science fiction writers – it is here, working its way steadily into countless aspects of modern life and the world of banking is no exception.
A wave of tech-savvy start-ups have been hard at work finding ways to introduce the technology into the banking process – from customer-facing chatbots to anti-money laundering systems.
But it isn’t just the fleet-footed fintech innovators that are jumping aboard the AI train – heavy hitters are also looking at ways to harness its power to strengthen their businesses.
Here we take a look at a couple of ways intelligent automation is being integrated into the financial system and companies looking to capitalise upon its potential.
How will AI banking trends affect the industry?
Artificial intelligence can be deployed by businesses in a number of different ways, but Matt Phillips, vice-president and head of financial services at retail and financial tech firm Diebold Nixdorf, suggests there are three key categories where AI is going to make a big difference.
He says: “The introduction of AI in banking doesn’t necessarily mean you’ll be seeing flashy robots zooming around your local branch any time soon. Instead, this is about a much subtler change.
“By driving the financial services industry to utilise machine learning and cognitive development, AI is affecting banking on three levels.
“Firstly, by improving what goes on behind the scenes; secondly, by allowing the innovation necessary to improve the customer experience; and thirdly by boosting security.”
AI-powered virtual banking assistants
Probably the most conspicuous way AI has made its way into the financial services industry is through virtual assistants providing customer service support.
While holographic bank tellers might not have yet arrived on the scene, the technology is nevertheless becoming gradually integrated into the ways customers interact with financial services.
Bank of America’s Erica
The Bank of America has had success with its roll-out of Erica, an app-based digital helper which attracted four million users in the six months after it was launched in May last year.
Customers are able to use voice, text and gesture commands to speak to Erica, which uses predictive analytics and natural language processing to simulate a real-life conversation with users when they ask questions or give instructions.
Erica is able to help with queries including past transactions, bill payments and locking or unlocking debit cards – and machine learning means each customer interaction with the assistant is logged and then used to improve future communication.
Bank of America’s head of consumer and wealth management technology, Aditya Bhasin, says: “Erica’s knowledge of banking and financial services increases with every client interaction.
“In time, Erica will have the insights to not only help pay a friend or list your transactions at a specific merchant, but also help you make better financial decisions by analysing your habits and providing guidance.”
IPsoft’s Amelia
US technology company IPsoft specialises in AI development and has also dipped its toe into the banking industry through own virtual assistant, Amelia.
Amelia was developed as a cognitive “digital colleague” that can be deployed visually and engages in natural, non-linear conversations with customers to help them solve queries.
Swedish bank SEB has used Amelia in a customer-facing role on its website to give customers round-the-clock access to an assistant that can help guide them through administrative processes that would otherwise require a branch visit or phone call.
IPsoft‘s executive director of transformations Johan Toll says: “Amelia’s natural language understanding and contextual understanding enables her to interact with humans to achieve their goal or intent.
“She is able to deal the with the inevitable lack of predictability and non-linear nature of real user conversations, resulting in a better understanding and less repetitive, clarifying questions – as well as removing unnecessary questions.
“With Amelia, SEB offers customers an opportunity to get an immediate response to their questions, which in turn concentrates the call volume for live human agents on the highest value support areas.
“Currently, Amelia is handling customer queries such as password resets for online banking accounts, helping users to step-by-step troubleshoot problems with credit and debit cards, and providing the location of the nearest bank.”
How banks are using AI in fraud prevention
Banks are also using artificial intelligence to improve their own internal process for analysing markets and assessing investment risks.
A key focus in this area is with anti-money laundering activities and the prevention of fraud.
Economic losses resulting from fraud a very high, with market analysis firm Crowe UK publishing research in 2018 that estimated the global cost of economic fraud, when taken as a proportion of global GDP for 2017, equated to £3.24tn.
Mimiro
One company working with AI in this field is Mimiro – a UK fintech start-up that recently raised £22.9m in financing for its anti-money laundering platform.
Mimiro uses machine learning technology to help its 350-strong client base make fast, intelligent risk decisions, as well as automating certain compliance and risk processes.
CEO and founder Charles Delingpole says: “We exist because globalisation is intensifying the business problems of trust.
“To offset concerns, many businesses can be hyper-cautious and conservative, losing out on commercial opportunities – in some cases abandoning entire countries or industries.”
The platform uses its AI capabilities to analyse streams of data with self-improving algorithms, identifying patterns that could signify fraudulent activity.
Data sources scanned by the AI include registers of high-level national and international sanctions, individuals who should be treated with caution, and adverse media coverage.
Pelican
Another company working with AI to combat financial crime is London-based tech firm Pelican, which supplies banks with tools to better detect and prevent instances of money laundering or fraud.
Pelican has a 20-year history of working with businesses in this way, the most up-to-date result being its PelicanSecure product suite.
PelicanSecure brings together tools that use natural language processing and machine learning to analyse patterns of behaviour to flag up “subtle anomalies” pointing to instances of fraud.
Factors like user location, spending patterns and unusual device configuration are all integrated into Pelican’s detection system.
Describing the role AI has to play in turning the tide against scammers, CEO Parth Desai says: “Traditional fraud detection methods are reactive in nature, meaning if a fraudster came up with a new idea to defraud an organization, the existing rules will fail to prevent it.
“AI, on the other hand, predicts those behaviours and protects against trending and future fraud typologies.
“AI, including machine learning, is unquestionably the future of fraud detection. Financial institutions are shifting towards AI gradually.
“We believe the industry is still in the launching phase of this technology and that there is a lot more to explore over the course of the coming few years.”
Other ways AI is being used in banking
The rewards of getting artificial intelligence implementation right are potentially huge – lower costs, increased security and higher customer satisfaction – and, accordingly, financial service providers are looking at a number of different ways to implement the technology.
Machine learning capabilities and the sheer data-handling power of AI are giving firms the chance to simplify and streamline multiple processes – for consumers and themselves.
In payments, for example, products like Apple Pay and China’s AliPay are using biometric scanning to enable users to authenticate digital transactions with their fingerprints and faces.
The process of paying for goods is one that has evolved significantly over time – from cheques, to chip and pin, to contactless – and AI is now taking this to the next level.
Meanwhile, personal finance and credit scoring apps such as Chip and Aire are making use of data aggregation and predictive analytics to assess customer finances and encourage better decision-making from both lenders and individuals.
Whatever the goal, the banking trend towards AI is rapidly gathering pace and looks set to continue to drive innovation within the industry.
