Five Application Scenarios of AI in Banking

DENIS OSTAPCHENYA | April 13, 2021

article image
Over the past decades, banks have been improving their ways of interacting with customers. They have tailored modern technology to the specific character of their work. For example, in the 1960s, the first ATMs appeared, and ten years later, there were already cards for payment. At the beginning of our century, users learned about round-the-clock online banking, and in 2010, they heard about mobile banking. But the development of the financial system didn’t stop there, as the digital age is opening up new opportunities — the use of Artificial Intelligence. By 2023, banks are projected to save $447 billion by applying AI apps. We will tell you how financial institutions are making use of this technology in their operations today. 

AI-powered chatbots
Chatbots are AI-enabled conversational interfaces. This is one of the most popular cases of applying AI in banking. Bots communicate with thousands of customers on behalf of the bank without requiring large expenses. Researchers have estimated that financial institutions save four minutes for each communication that the chatbot handles.
Since customers use mobile apps to carry out monetary transactions, banks embed chatbot services in them. This makes it possible to attract users’ attention and create a brand that is recognizable in the market.
For example, Bank of America launched a chatbot that sends users notifications, informs them about their balances, makes recommendations for saving money, provides updates to credit reports, and so on. This is the way the bank helps its clients to make informed decisions.
Another example is the launch of the Ceba chatbot, which brought great success to the Australian Commonwealth Bank. With its help, about half a million customers were able to solve more than two hundred banking issues: activate their cards, check account balances, withdraw cash, etc.

Mobile banking
AI functionality in mobile apps is becoming more proactive, personalized, and advanced. For example, Royal Bank of Canada has included Siri in its iOS app. Now, to send money to another card, it’s enough to say something like: "Hey, Siri, send $30 to Lisa!" - and confirm the transaction using Touch ID.
Thanks to AI, banks generate 66% more revenue from mobile banking users than when customers visit branches. Banking organizations are paying close attention to this technology to improve their quality of services and remain competitive in the market.

Data collection and analysis
Banking institutions record millions of business transactions every day. The volume of information generated by banks is enormous, so its collection and registration turn into an overwhelming task for employees. Structuring and recording this data is impossible until there is a plan for its use. Therefore, determining the relationship between the collected data is challenging, especially when a bank has thousands of clients.
There used to be the following approach: a client came to a meeting with a bank employee who knew their name and financial history and understood what options were better to offer. But that's history now. With the wealth of data coming from countless transactions, banks are trying to implement innovative business ideas and risk management solutions.
AI-based apps collect and analyze data. This improves the user experience. The information can be used for granting loans or detecting fraud. Companies that estimated their profit from Big Data analysis have reported an average increase in revenue by 8% and a reduction in costs by 10%.

Risk management
Extension of credit is quite a challenging task for bankers. If a bank gives money to insolvent customers, it can get into difficulties. If a borrower loses a stable income, this leads to default. According to statistics, in 2020, credit card delinquencies in the U.S. rose by 1.4% within six months.
AI-powered systems can appraise customer credit histories more accurately to avoid this level of default. Mobile banking apps track financial transactions and analyze user data. This helps banks anticipate the risks associated with issuing loans, such as customer insolvency or the threat of fraud.

Data security
According to the Federal Trade Commission report for 2020, credit card fraud is the most common type of personal data theft.
AI-based systems are effective against malefactors. The programs analyze customer behavior, location, and financial habits and trigger a security mechanism if they detect any unusual activity. ABI Research estimates that spending on AI and cybersecurity analytics will amount to $96 billion by the end of 2021.
Amazon has already acquired harvest.AI - an AI cyber security startup - and launched Macie - a service that applies Machine Learning to detect, sort, and structure data in S3 cloud storage.

Spotlight

Cembra Money Bank AG

Cembra Money Bank offers personal loans, vehicle finance, credit cards, savings products and insurance. It is committed to the principles of maximum transparency, long-term advice and the greatest possible degree of respect towards all its stakeholders.

OTHER ARTICLES
CORE BANKING

Open Banking and Trust a key issue

Article | April 15, 2021

Open Banking is all about the customer being in control of their data and funds. It gives them the freedom and flexibility to decide when and with whom to share their valuable information. However, as with all vibrant and progressive ecosystems, speed, security, and ease of use will determine open banking’s future success along with the key issue of trust. Will the end user trust people to share data with them and trust their banks to still protect their data? PSD2 Open Banking gives Payment Service Users (PSUs) the legal right to share their transactional account data with regulated third party providers (TPPs). For this to be possible, the 6,000+ Financial Institutions providing transactional payment accounts that can be accessed online have to put in place open banking APIs. These APIs give TPPs the access required to either make payments on an account holder’s behalf or view account data and funds, both of which require the account holder’s prior explicit consent. Access can only be denied if a TPP is believed to be unauthorised or fraudulent. Open banking regulation has given rise to a new group of FinTechs who are seizing the opportunity to create innovative apps and products with the customer at the core of the offering. At the end of 2019, 240 TPPs from across the EEA and UK were regulated to provide open banking services. A year later, this figure had increased to 450 (excluding the thousands of credit institutions that are also able to act in the capacity of TPPs). The near doubling of newly regulated entities demonstrates user demand for the innovative products and services that these organisations are offering – it is now down to trust and security in the ecosystem, along with ease of use, to drive volumes. Source: Konsentus The ability for TPPs, many of whom may be unknown to these Financial Institutions, to request immediate access to valuable data and funds presents many challenges and risks – all of which must be addressed without introducing potential friction in the customer journey. The main challenges are knowing if a TPP is who it claims to be and whether it is regulated to provide the services being requested at the time of the transaction request. After all, these are the key factors enabling the bank to trust the TPP and feel confident the end user can trust them. The added difficulty of knowing which markets within the EEA a TPP is authorised to operate in is an additional challenge. Financial Institutions have long been the trusted guardians of their customers’ data and funds. Although the open banking model means the customer now has ultimate control of their data, it is still primarily the Financial Institution’s responsibility to ensure nothing goes wrong and they are likely to be held liable in any disputes that arise. There is also the very real reputational risk to Financial Institution if something does go wrong. Checking a TPP’s identity, its current regulated status, and the services it is requesting to perform are essential but not easy tasks to complete in that, firstly, a Financial Institution needs to determine whether a TPP is who it claims to be. This is done by having real-time access to the 70+ Qualified Trust Service Providers (QTSPs) who can issue PSD2 eIDAS certificates. These eIDAS certificates contain the requisite information on a TPP’s identity and are used to secure communications between Financial Institutions and TPPs. They also digitally seal messages, ensuring the integrity of the concept and proof of origin. However, an eIDAS certificate can have up to a two-year validity period. During this time, changes may have been made to a TPP’s regulatory authorisation status by its Home National Competent Authority (NCA). This introduces significant risk to the Financial Institution’s decision process. eIDAS certificates also do not contain information on the countries a TPP is authorised to provide their products and services into under passporting rules. This information is held on the TPP’s Home NCA Credit Institution and Payment Service Provider (PSP) registers. Between them, the 31 NCAs maintain over 115 databases and registers. Checking them at the time of a transaction request is paramount to prevent fraudulent TPPs from slipping through the net. According to the Konsentus Q4 2020 TPP tracker, every country in the EEA had at least 75 TPPs who could provide open banking services. These may not all be Home regulated TPPs. Take, for instance, Germany, who had 35 Home Regulated TPPs in December 2020 but an additional 112 TPPs who could passport in their services. To do the requisite due diligence on all these TPPs would require having online access to all the databases and registers hosted by the NCAs regulating these TPPs. This means connecting to the 31 NCAs and interrogating over 115 separate registers in real-time, in addition to connecting with all the QTSPs who issue PSD2 eIDAS certificates. When a Financial Institution is presented with an eIDAS certificate by a TPP, if a real-time online connection can be made to all the legal sources of record, the Financial Institution can make an instant informed risk management decision on whether, or not, to give the TPP access. All this can be done behind the scenes without the end user even being aware of what is happening. As volumes look to dramatically increase over the next few years fraudulent and other sorts of attacks are bound to increase. Financial institutions are going to face increasing challenges around protecting end users’ data, ensuring access is only given to those with the appropriate authorisations and permissions. A very real risk for them is the reputational one; after all, end users may not be that good at separating a reputational issue around open banking from broader issues around their banking relationship. For Financial Institutions, maintaining trust in their brands is going to be crucial going forward, but the risks are going to increase if they have not locked down who can access end user account data and funds.

Read More

How Financial Advisors Can Use Social Media to Build a Brand

Article | April 15, 2021

Most advisors know they should be using social media, but running a successful advisory business requires dealing with a lot of moving parts. When you’re handling a client crisis or slogging through the daily routine that comes with managing a team of financial professionals, it can be easy to let something like posting on Twitter fall by the wayside. After all, isn’t social media just a distraction from your actual work?

Read More

How machine learning is changing financial services

Article | April 15, 2021

Artificial intelligence (AI) has become integrated into our everyday lives. It powers what we see in our social media newsfeeds, activates facial recognition (to unlock our smartphones), and even suggests music for us to listen to. Machine learning, a subset of AI, is progressively integrating into our everyday and changing how we live and make decisions.

Read More

Does robotic process automation have a place in finance?

Article | April 15, 2021

Robotic process automation (RPA) may be a relatively young technology, but it’s already secured a firm foothold in the financial services industry. In the UK alone, 72% of banks, insurance firms and other financial institutions are using artificial intelligence (AI) technology, according to Microsoft’s report, ‘Accelerating competitive advantage with AI’.

Read More

Spotlight

Cembra Money Bank AG

Cembra Money Bank offers personal loans, vehicle finance, credit cards, savings products and insurance. It is committed to the principles of maximum transparency, long-term advice and the greatest possible degree of respect towards all its stakeholders.

Events