How to make use of the potential of AI in banking? Part 1.
The dynamic growth in access to structured and non-structured data, the availability of new technologies, such as cloud computing and machine learning algorithms, the growing pressure caused by new competition, the increase in regulation, and the increased consumer expectations – all this make implementation of artificial intelligence (AI) as one of the most pressing issues for companies from the banking sector. So how AI in banking is implemented?
There are no doubts that algorithms based on machine learning will change the character of financial corporations. Facing new technological challenges becomes not only an element of creating advantage over competitors but, above all, a foundation allowing the company to stay on the market. Meanwhile, the competition is persistent. Since 2000, more than a half of financial companies from the Fortune 500 list have failed to endure the intense competition and have dropped out of the business game.
Artificial intelligence certainly makes competition between companies from the banking sector enter a completely new – more intense – level. At the same time, however, this technology brings along tremendous possibilities. Despite the fact that AI is still at its development stage, it is already reflected in real business performance for many companies. Artificial intelligence is currently something more than just an element that improves efficiency – it’s a completely new production factor. Its effect on business may be revolutionary. Intelligent automation lets us almost completely eliminate laborious, repeating tasks from employees’ lives, increase effectiveness and efficiency of their decisions, improve relations between the consumer and the company, and additionally better allocate and invest the resources. How does the use of artificial intelligence in banking looks like?
Possibilities of application of AI in banking
Artificial intelligence will have a crucial impact on the development of financial services in the years to come. Algorithms based on machine learning will re-define the way banks work (their processes), what they sell (their products and services) and how they communicate with their clients (user experience). People and AI systems will work together, increasing efficiency and improving experiences of both consumers and employees.
Improvement of processes:
Today, many banks already have systems based on AI which take over execution of low-value processes. It means that all processes that require particular analysis done by humans are carried out automatically (e.g. any kind of documents can be scanned and analyzed by computers). The potential brought by the use of artificial intelligence in process improvement is almost infinite and the spectrum of technologies used in this respect is very broad.
For example, image recognition and machine learning can be combined to scan masses of documents and take actions on the basis of the valid regulations. Algorithms can then be used to decide which cases should be referred to a person making financial decisions, and which may be analyzed completely by a machine.
Another example is the use of AI by a group of banks in Australia and New Zealand, where they implemented almost complete automatization of more than 40 various processes, including repayment of mortgage credits and creation of half-yearly audit, containing data collected from over ten various systems. Currently, thanks to the implemented innovations, the bank gains savings in some areas even amounting to more than 30 percent (annually). Furthermore, a new system allows employees to focus on tasks that are definitely more satisfactory and have a higher value both for them and for an organization.
Barclays, a British holding company and global provider of financial services operating on all continents, also introduced automation of a number of processes, such as analysis of receivables and verification of unfair practices, thus reducing reserves intended for securing bad credits by about USD 225 million per year.
The potential brought by artificial intelligence is often – in the context of banking – associated with the possibility of performing a detailed analysis and profiling of each bank client.
Indeed, financial companies effectively utilize AI algorithms for identification of clients who will most likely leave the bank in the near future. For this purpose, they are using chatbots integrated with the existing messaging platforms and with interfaces of bank websites. Similar tools enriched by data, such as held assets or transaction history, are utilized in identifying clients who are or will in the near future search for possibilities of investing their funds.
Furthermore, machine learning is also a technology that completely changes the rules of the game for the creditors. It allows decision-makers to obtain a faster and more precise credit risk assessment for each incoming application, and it can also evaluate financial capacity of applicants on the current basis. Some banks already use dynamic credit risk models, extremely useful in determining creditworthiness of young people with a limited credit history or self-employed people. Artificial intelligence may also bring a number of benefits to consumers, e.g. help creditors in repaying outstanding debts.
Virtual client service:
There are many examples of companies that implement virtual client service based on AI. For example, JPMorgan uses robots to respond to internal IT inquiries, including resetting of passwords for employees (and soon also for clients). Bots in this bank are to handle 1.7 million requests of access to IT resources this year, performing the work of 40 full-time employees. Another example is Fukoku Mutual Life Insurance, a Japanese banking and insurance company, where the Watson Explorer system created by IBM will perform the work of 34 employees. Its task will involve analysis of claims received by the company. (More AI solutions in client service in the second part of the article link)
Fraud detection and risk management
Algorithms of artificial intelligence in banking can detect potential frauds, even before they actually take place. This technology can quickly imitate the thinking process of a real analyst and look through each transaction, in every portfolio within the bank’s resources. AI allows financial institutions not only to receive warnings about potential frauds, but also to perform accurate percentage identification of the likelihood of occurrence of embezzlements. Of course, it involves a number of benefits, both for the bank and for its clients.
AI helps banks replace only those cards that may really be the object of fraud – only a small percentage of cards threatened by a breach may actually become used for embezzlement. The possibility to identify specific accounts in danger allows banks to save a substantial amount on fees under re-issuance of payment cards, and also allows them to better allocate their funds to securing banking systems.
The above examples are but a few of the many possibilities of applying artificial intelligence in banking. The second part of the article is going to present several tips on how to prepare a company for implementation of systems based on AI, what to pay attention to and how the potential of artificial intelligence was used by 3 largest American banks.
Looking for the part 2 od this article? It’s here.