-
AI Payments

AI Payments Revolution

26 April 2019

Transaction banking is an area that provides a perfect number of conditions for AI to flourish. It is characterised by a large number of processes, supporting large number of customer interactions with the bank for day to day banking, generating large volumes of data; and that potentially makes transaction banking the perfect playground for AI.

Artificial intelligence (AI) is a revolution that will completely transform the business landscape over the next decade. This set of technologies, which is capable of carrying out tasks that previously required human intelligence, is going to change how businesses interact with their customers, operate their machines, use their people, develop products and services, and plan for the future.

Examples of AI technologies include cognitive reasoning (replication of the human judgment process), deep learning (learning by example), machine learning (learning from experience) and natural language processing (drawing conclusions from natural language data). These technologies can be particularly powerful when they are used in conjunction with robotic process automation (RPA), which automates mundane tasks. AI can extract valuable information and identify patterns from the high-quality, structured data that is generated by RPA.

AI is already being applied to a wide range of business activities, from farming shrimps and brewing beer through to building self-driving cars and optimising oil and gas production. The financial services sector is a pioneer in the adoption of AI, applying the technologies in multiple areas, from fraud detection and conduct surveillance through to stock picking. Chatbots are used to complement banks’ customer service offerings and there is huge potential for AI to be adopted widely within transaction banking in the future.

AI and transaction banking

Transaction banking is characterised by a large number of processes, supporting a large number of customer interactions for day to day banking, generating large volumes of data; and that potentially makes transaction banking an excellent use case for AI. In particular, AI presents significant opportunities in areas such as fraud detection and prevention, payments and on-boarding.

AI can help to prevent and detect fraud by flagging up unusual transactions – for example, where the amount involved is very large, or the transaction was initiated by someone unexpected, or where the organisation has never previously transacted with the destination company or country that is receiving the payment. Additionally, AI tools can detect and monitor unusual behaviours in staff, such as logging on to banking systems out of hours.

Another important use case is payments. AI can be used to improve the speed and efficiency of the payment process, by reducing the extent to which humans need to be involved. For example, today the process of paying a simple invoice can involve significant human intervention both for the corporates and their bank, but AI can facilitate straight-through processing of payments, by automating workflows, providing decision support and applying image recognition to documents. Also, developments in speech recognition technology mean that banks can increasingly process payments initiated via voice, where the initiator has used a smart phone or smart speaker. 

Banks’ procedures for on-boarding new corporate clients could be much smoother in future if they take advantage of AI technologies to process the vast swathes of documentation required for Know Your Customer (KYC) purposes. Machines will be able to use natural language processing to read through the documents, make sense of them and feed back their findings to human decision makers. They will also be able to cross-reference the documents with external sources, such as Companies House. The result should be a faster and more efficient on-boarding process that is a much better experience for both banks and their clients.

Regulatory compliance is another area where AI is likely to have an impact on transaction banking. The technology could be used to validate client transactions against money laundering and sanctions rules and detect patterns that indicate illegal activity. Whilst there is no evidence of widespread usage of AI for this purpose yet, it is likely to come in future.

Although there are many potential use cases for AI within transaction banking, customer service is the area where it is probably applied most today. Chatbots (computer programs that are designed to conduct conversations with humans using voice or text) are already able to respond to simple queries from clients and carry out basic tasks, such as creating or cancelling a standing order or Direct Debit, or giving more information on a payment that the client doesn’t recognise. Some financial organisations use chatbots to carry out FX trades.

Overcoming the challenges

While transaction banking lends itself to the large-scale application of AI, adoption of the technologies rests on banks and their clients overcoming some key challenges:

  1. Data. Organisations are producing reams and reams of data that could potentially be interrogated by AI technologies. This data comes in both structured form (such as bank account reconciliations) and unstructured form (such as emails) and often, various pieces of data are not connected or appropriately valued. This poses a significant challenge. Furthermore, one machine cannot necessarily read data produced by another machine. Organisations are still producing data in vertical silos, rather than through an end-to-end horizontal process, which acts as a barrier to the adoption of AI.

  2. Culture. AI outcomes inspire a huge amount of fear and awe, partly as a result of the media coverage that they generate. If people within banks and client organisations are afraid of losing their jobs, they may be reluctant to embrace AI in a way that realises its full potential. Also some clients may feel uncomfortable about talking to a chatbot, or fear that in doing so, they might inadvertently expose their organisation to data privacy breaches.

  3. Technological limitations. While huge strides are being made in AI, the technologies still have some major limitations to overcome. For example, chatbots currently lack the cognitive ability to wrestle with complex challenges or human emotions, such as irritated customers – although this could change in future. Also, AI applications are only as good as the data they are trained on. If data is inaccurate or contains inherent bias, that will be reflected in the results generated by AI.

  4. Business models. To make optimal use of AI, banks and their clients will need to adopt new business models, procedures and behaviours. They must also have good processes for collecting, storing and exploiting data, and make use of supporting technologies, such as the cloud and application program interfaces or APIs, to facilitate smoother processing of transactions.

    User interfaces are another important consideration – there is no point in deploying state-of-the-art chatbots if a system’s existing customer interface is awkward, out-dated and slow. All round, companies need to ensure that, their systems and processes are set up to take full advantage of AI.

Where next?

AI is undoubtedly part of the future of transaction banking, although it’s not yet easy to predict exactly how widely it will be adopted. The huge volumes of data, interactions, processes and transactions involved with transaction banking make it ideally suited for application of the technologies. With AI, banks will gain from huge data-processing capabilities at a low cost while clients will enjoy improved security and an enhanced customer experience.

Of course, the extent to which the power of AI will be realised within transaction banking rests upon the creativity and skill of the organisations that deploy it. Today, AI is very much a work in progress, but if we can overcome the challenges associated with using the technologies, both banks and their clients will reap huge benefits. AI will be core to transaction banking in the future – we must be ready for it and ready to embrace it as the revolution becomes real!

*
Insight

Driverless Cars

The driverless car is no longer a distant vision of the future, but a real prospect for our roads and an opportunity for fresh economic growth.

*
Insight

Evolving Payments Landscape in Europe

This article takes a closer look at two recent regulatory updates that will impact the future of the European payments industry.