For financial institutions, recovering from the pandemic will put an end to tentative experiments with artificial intelligence (AI) and machine learning (ML), and demand their large-scale adoption. The crisis has required financial organisations to respond to customer needs around the clock. Many are therefore transforming with ever-increasing pace, but they must ensure that their core critical operations continue to run smoothly. This has sparked an interest in AI and ML solutions, which reduce the need for manual intervention in operations, significantly improve security and free up time for innovation. Reducing the time between the generation of an idea and it delivering value for the business, AI and ML promise long-term, strategic advantages for organisations.
So how can banks and finance institutions make the most of AI, and what are the key use cases in practice?
Many financial services firms had already adopted AI and ML prior to the pandemic. However, many had difficulties identifying which key functions benefit most from AI, and so the technology did not always deliver the returns expected. This is set to change in the coming months: increased AI and ML deployment will be at the heart of the economic recovery from COVID-19, and the pandemic has highlighted particular areas where AI should be applied. These range from informing credit decisions and preventing fraud, to improving the customer experience through frictionless, 24/7 interactions.
Some specific financial services processes that can be improved by AI include:
Intelligent and robotic process automation optimise various functions, enhance efficiency, and improve the overall speed and accuracy of core financial processes, leading to substantial cost-savings. One area that has risen in prominence is e-KYC, or ‘electronic know-your-customer’. This is a remote, paperless process that reduces the bureaucratic costs of crucial ‘know-your-customer’ protocols, such as verification of client identities and signatures.
This task once involved repetitive, mundane actions with considerable effort required just to keep track of document handling, loan disbursement and repayment, as well as regulatory reporting of the entire process. However, this year, organisations are embracing intelligent automation platforms that manage, interpret and extract unstructured data, including text, images, scanned documents (handwritten and electronic), faxes, and web content. Running on an NLP (natural language processing) engine, which identifies any missing, unseen, and ill-formed data, these platforms offer near-perfect accuracy and higher reliability. Average handling time is reduced, and firms gain a significant competitive advantage through an improved customer experience.
Virtual assistants can respond to customer needs with minimal employee input. A straightforward means of increasing productivity, the time and effort spent on generic customer queries is reduced, freeing up teams to focus on longer-term projects that drive innovation across the business.
We’re all familiar with chatbots on e-commerce sites, and such solutions will become increasingly common in the financial services industry, with organisations such as JP Morgan now making use of these bots to streamline their back-office operations and strengthen customer support. The platforms involve COIN, short for ‘contract intelligence’, which runs on an ML system powered by the bank’s private cloud network. As well as creating appropriate responses to general queries, COIN automates legal filing tasks, reviews documents, handles basic IT requests such as password resets, and creates new tools for both bankers and clients with greater proficiency and less human error.
Estimating creditworthiness is largely based on the likelihood of an individual or business repaying a loan. Determining the chances of default underpins the risk management processes at all lending organisations. Even with impeccable data, assessing this has its difficulties, as some individuals and organisations can be untruthful about their ability to pay their loans back.
To combat this, companies such as Lenddo and ZestFinance are using AI for risk assessment, and to determine an individual’s creditworthiness. Credit bureaus such as Equifax also use AI, ML and advanced data and analytical tools to analyse alternate sources in the evaluation of risk, and gain customer insight in the process.
Lenders once used a limited set of data, such as annual salaries and credit scores, for this process. However, thanks to AI, organisations are now able to consider an individual’s entire digital financial footprint to determine the likelihood of default. In addition to traditional data sets, the analysis of this alternative data is particularly useful in determining the creditworthiness of individuals without conventional records of loan or credit history.
The way that businesses and clients interact with each other has changed irreversibly this year, and the finance industry is no different. Before the urgency demanded by the pandemic, financial institutions had been experimenting with AI and ML on a limited scale – mainly as a tick-box exercise in an effort to ‘keep up with the Joneses’. The widespread adoption that has been taking place this year stems from the need to truly innovate and increase resilience across the sector.
Banks and financial institutions are now aware of the key areas that benefit from AI, such as greater efficiency in back office operations, and significant improvements in customer engagement. A transformation process that was in its infancy prior to Covid-19 has accelerated and is fast becoming the standard approach. What’s more, financial organisations that are embracing AI now and prioritising its full implementation will be best placed to reap its rewards in the future.
(Source: AI News https://artificialintelligence-news.com/2020/12/15/from-experimentation-to-implementation-how-ai-is-proving-its-worth-in-financial-services/ )
Chinese
Japanese
Korean
