Banks and users are experiencing a digital commerce acceleration, promoted greatly by the time we’re spending at home. In an online forum, we looked at the role fintechs play and the importance of trust in fintech, from fraud prevention to managing credit risk.
At the recent Mastercard® Fintech Virtual Forum, Sudhir Jha, Mastercard Senior Vice President and Head of Brighterion, spoke about AI for decision making and announced the expanded credit risk solution, the Brighterion AI Credit Risk Suite. Jha was joined by Mohammed Abdelsadek, Mastercard’s Executive Vice President of Cyber and Intelligence, to discuss the role of trust in digital financial services. This insightful panel discussion was moderated by Diego Szteinhendler, SVP of Cyber and Intelligence at Mastercard. In this post, we’re exploring the highlights of their discussion.
Buildin g and maintaining trust in digital financial services
The digital user journey begins at account creation and continues throughout the customer experience, from the initiation of a transaction to completed payment. Brighterion plays a key role in Mastercard’s fraud prevention network with its Decision Intelligence technology, including authentication of users and verification of transactions.
S
zteinhendler laid the groundwork for the discussion: “Clearly consumers are enjoying the benefits of digital commerce acceleration. But as you use different platforms, there’s also a fear of the unknown and security, and how these platforms are protecting consumers. The key thing when consumers use these platforms is trust and how to enable trust.”
The first opportunity to establish trust in fintech is when creating a new account. Getting that right the first time is the basis and the foundation of your relationship with your new customer. The second is maintaining that trust in ongoing visits or transactions. The last is making sure that we share among various stakeholders the data that is used to create that trust and create frictionless decisions to maintain that trust and to use the data for accuracy and predictability.
Low trust comes with high cost
“Trust is one of the most important aspects in digital financial services. The moment you lose trust, it becomes difficult to re-establish, which has an impact on businesses and consumers,” says Abdelsadek. “A recent study by Juniper Research found that the cumulative losses from fraud, ecommerce, banks, and e-transfers will be over $200 billion over the years 2020 to 2024. That is a significant number.”
Most customers expect some sort of friction to ensure their transactions are secure, whether it’s a CAPTCHA riddle, answering a security question, or entering a security code sent to another device. Financial institutions (FIs) and fintech are equally conscious of the delicate balance between friction and a positive customer experience. When friction is too low, customers worry their transactions aren’t secure. When it’s too high, friction can cause customers to abandon transactions and leave their FIs for ones with friendlier platforms. These factors build – or diminish – trust.
Fraud costs significant reputational damage, too. A 2019 study conducted by Bing Identity found that 81 percent of the respondents would stop engaging with a brand or line following a data breach, Abdelsadek says. This lost trust is very difficult and expensive to regain.
Machine learning is key to AI decision making
Machine learning models depend on data and the more data fed into the system, the more accurate the decision. More data means less friction.
AI decisioning works to determine authenticity of customer identity by collecting data with each interaction, for example. The dilemma is to not create excessive friction while collecting the data for the initial and future authentications.
“There are two broad techniques that we use to do that. One is that we mimic a physical interaction in a digital environment,” says Jha. “For example, if you go to a neighborhood store for the first time, the store owner is not necessarily going to know you and therefore would ask a few more questions, would try to establish your behavior, your mannerisms, your preferences, and then create a mental model of you in their mind.”
Jha explains that AI is similar. “The idea is just like the first time you interact with a store or do a transaction, there’s going to be a little bit more data collection, a little bit more information that needs to be input. But subsequently, similar to the neighborhood store where the person will get to know you better and ask fewer questions, digital interactions will involve less friction.”
Through machine learning, Brighterion’s AI learns the customer’s online patterns with each transaction. It can ask fewer questions in subsequent interactions and make intelligent decisions based on the accumulated data.
Advanced AI learns and recognizes your customers
The second technique in recognizing your customers involves implicit data collection. Data collected by asking questions isn’t enough to stop fraud; fraudsters can often easily answer those questions. Advanced AI can learn each customer’s IP, their preferred language, and the type of device they use and how they hold it. These behavioral and other implicit data can be collected to further inform machine learning models.
With these two cornerstones, Brighterion AI allows normal transactions to proceed without friction. When there’s a suspicious transaction, the user may be asked a follow up question instead of being denied. If the question is answered accurately, the transaction proceeds.
When trusted customers change behavior: managing credit risk
In a year when the economy has been hit hard by a global pandemic and other systemic factors, Mastercard believes there is a role to play in helping consumers and small-to-medium-sized businesses rebuild. Jha acknowledges it will be important to manage the credit risk portfolio, which led to the Brighterion AI Credit Risk Suite. Expanding the previous credit risk solution, this broader solution encompasses credit origination, portfolio management and delinquency, and collections optimization.
“The idea is to provide the credit, manage the credit, predict the delinquency, and collect on the credit,” Jha says. “At the same time, using this suite of credit risk tools ensures you’re providing the maximum flexibility to consumers. I believe 2021 is going to be that year where a lot of businesses, a lot of users, are going to be trying to build their lives back. The more financial institutions that participate, and AI models can definitely help, the better and faster we can recover.”
The Brighterion AI Credit Risk Suite
Brighterion’s AI decisioning tool plays a strong role in the Brighterion AI Credit Risk Suite. Starting at credit origination, the solution helps lenders establish trust by verifying the user’s data, ensuring applications aren’t fraudulent and determining the customer’s credit worthiness. The AI model can help lenders and borrowers build more robust applications, especially for thin file applicants. The solution provides a very good approximation of credit limits, which can be adjusted up or down over the customer lifecycle.
Once approved, the account moves into portfolio management. FIs will use AI to assess customers and help decide if they are eligible for additional credits, rewards or new products. It will also identify users who are unable to manage their credit and may be headed for trouble.
The credit risk solution’s AI model predicts credit delinquency as early as 120 days in advance, allowing lenders to take collective action. Helping customers by offering payment plans, temporarily reduced interest and other actions will help to prevent millions of dollars in credit loss while increasing customer loyalty and building trust.
If the delinquent account can’t be resolved, it may go to collection. The collection optimization feature allows lenders to make informed decisions. The Brighterion AI Credit Risk Suite will segment collections to ensure FIs use the best strategy for each borrower.
AI is proving to be a powerful tool in fintech. Mastercard revolutionized the use of AI in payments over the past 15 years, offering innovative solutions to the financial services industry. From its use of Brighterion’s Decision Intelligence that protects customers along the digital journey to their timely Credit Risk Suite that enables borrowers to access and maintain good credit, Mastercard helps reduce risk for issuers and lenders.
Want more? View the on-demand recording of the session Building trust in AI throughout the digital user journey here.
