Artificial intelligence (AI) is seemingly everywhere these days, from Siri on an iPhone to calling an Uber or watching that video recommended on Netflix based on predictive algorithms. While AI hasn’t quite taken over just yet, it’s clear that it will become even more deeply embedded in our lives. The banking industry, as much as any other, has the potential to proactively harness AI technology to transform itself, or use it to just keep up or get left behind.
With the market changing rapidly not only due to advanced digital technologies, but also emerging competition from fintechs and more knowledgeable, demanding customers, banks are faced with a number of challenges. To maintain market share and profitability in an industry where competing products and services are frequently very similar to their own, banks need to focus on what improves the customer experience to distinguish them from their peers and proactively service their customers. How they adopt and utilise AI is clearly going to be key.
Dipping a toe in the water
As a traditionally conservative industry, banking is slow to change. But the imperative for banks to differentiate themselves has forced them to look at how they can manage and predict customer behaviours. AI and machine learning give banks the opportunity to stand out from the pack by the offering them the ability to offer behaviour-based, predictive, more personalised customer services than ever before, and an easy way to do this at scale.
Another aspect to this is that today’s users and consumers are changing too. Whether commercial or domestic users, internal managers or staff, they are more open to having a hybrid human/AI-supported interface. This is becoming prevalent due to smartphone usage and widespread market consumerism. A recent Genpact survey found that 27 per cent of retail banking customers would be comfortable setting up a new bank account with a digital assistant, such as Alexa or Siri. This presents an opportunity, but also a threat, to non-banking lenders such as PayPal, Facebook and Amazon, who have significant brand awareness and access to capital and technology. It is clear that these businesses have the potential to disrupt traditional banking models. Without harnessing the power of data and AI more effectively, banks will not be able to compete with these firms.
In addition, after significant investment in robotic process automation (RPA), banks are discovering that many of the processes to which they have applied RPA will have a greater transformational impact and drive greater change when they add cognitive AI technology to the automation of these processes.
Deploying digital applications with success
While banks recognise the need to start utilising AI in their business operations, adoption varies. Most, if not all, banks have AI in a trial or in incubation at this point. One of the most intense areas of interest is around conversational AI, with organisations implementing chatbots to assist customers online and support contact centres. In commercial banking, we’re also seeing AI applications with loan origination and management supporting financial spreading and credit decisioning, where the ability to accelerate decisioning gives improved conversations and opportunities for risk reduction and enhanced capital efficiency, along with the traditional efficiency play.
We see that banks are already very comfortable with implementing conversational AI technologies, such as chatbots supplemented by an element of human support, because customers have shown they are ready to embrace these new technologies, especially where they do not provide financial transaction features. Genpact research suggests that customers would welcome a digital banking experience, with many saying that online solutions such as chatbots within mobile apps make banking easier. However, there is a degree of discomfort in using AI with transactional processing, a process whereby cognitive processing will make recommendations based on multiple data sources and data types such as unstructured data and apply prediction. Organisations are rightly cautious about this and are focused on questions of how to set policies and manage risk and compliance as they implement AI-powered technologies. The same argument applies to driverless car technology, where the case is made that an AI-powered car will be safer than a human driver.
To be successful in applying AI technology, banks must always consider the full customer experience – not just the upfront experience, but how they could apply AI across the journey through the middle and back office, and then how that could be used to offer new solutions to deliver the best service possible. This is the way in which banks can differentiate themselves in the market. Many banks have not made the most of AI’s potential. Too often they focus more on the front end, or selected restricted use cases based on current operating models, without sufficiently exploiting integrated technology and processes throughout the enterprise to transform the customer experience.
Looking over the horizon
With a growing level of comfort and an increasing amount of experience in implementing AI, we see three broad areas where banks could and should look to apply AI in the future. First, in the front end, AI can be expanded to the channels to deliver a more natural engagement layer experience, which is applicable to retail and commercial banking segments. There is enormous potential in implementing chatbots and voice AI in the omnichannel customer experience, a large element of which lies in at the interface.
Next, there are significant opportunities for having the capability to offer value-added personalisation services at scale. Banks could harness transactional histories, behavioural data and machine learning predictive algorithms to offer the personalisation currently offered by branches, contact centres and relationship managers. The more advanced banks will also look to work with non-banking partners or fintechs as they redefine the types of services that can be offered.
Finally, AI will transform back office functions, such as how banks manage risk, fraud and capital. We are already seeing AI applied to process external structured and unstructured data and news feeds. This technology can analyse multiple feeds to assess the potential impact on banks’ clients’ businesses based on materiality, as well as flag potential capital risks of client portfolios. This type of algorithm will, over time, learn with greater accuracy and give banks improved ability to forecast. The same AI components can equally be applied to fraud and anti-money laundering detection, particularly helping reduce the volume of false positives. In this sense, this highlights the ability of AI to deal with materiality and relationships of information when processing large data sets as an important element of machine learning and data analytic capabilities.
With personal and commercial banking tech-savvy customers ready to embrace new digital technologies, banks’ opportunity to use AI as a key differentiator in the market is stronger than ever. The firms that significantly embrace and exploit AI, machine learning and other relevant technologies will have a head start on the pack to improve cost and income ratios, and in the long term will effectively be the ones that stay in the lead.
Nick Lincoln, Partner & Europe Digital Consulting Head, Banking, Financial Services & Insurance, Genpact
Image Credit: Razum / Shutterstock