Online banking

How AI and ML Can Help Secure Online Banking

Financial services have become increasingly digital in recent years as financial institutions (FIs) and FinTechs have sought to meet increasing customer demands for convenience and speed. Widespread changes by consumers to stay at home during the pandemic have made remote financial services a practical necessity, with a survey last year finding that 73% of American adults were more likely to use banking services and digital payments while socially distancing.

Financial services providers who make the inevitable leap, however, must strike a balance between making these offers easy for customers to integrate and protecting them from fraudsters and cybercriminals. Transparent and robust identity verification can be a challenge to achieve, but the stakes are high: FIs that fail to detect fraudulent users and allow them to open accounts can open the door to money laundering. money, terrorist financing and other crimes. Such activities have a huge impact, with a 2020 report estimating that $ 800-2 trillion is laundered globally in an average year. This has prompted more governments to adopt initiatives requiring compliance with anti-money laundering (AML) efforts, requiring FIs to provide strong identification measures to weed out fraudsters. Customers are also concerned about the increase in fraud. Seventy-six percent of Canadian consumers in a recent survey said the mortgage industry needs to do more to protect their data, for example, and 40 percent fear fraudsters could steal their personal data and take out loans on their behalf.

These concerns are compelling, but complex and slow authentication processes that can distort the identification of legitimate users as fraudulent can result in business losses for FIs. A 2020 study found that 32% of UK-based consumers would abandon loan applications if the whole process couldn’t be completed in one channel, for example. Clients who are asked to complete online applications by verifying their identity in apps or scanning and emailing information may decide to ditch lenders in favor of competitors who offer more streamlined approaches. This in-depth dive examines how some financial services companies are refining their approaches with advanced technologies to ensure efficient and accurate identity verification.

Authentication challenges posed by AML compliance

Onboarding time is a major challenge in AML compliance. Many clients want to be approved within 24 hours, and overly complicated requests for personal information can lengthen the process, leading to potential frustration and abandonment. So collecting the right details quickly is essential, but many FIs struggle to achieve this. Companies working with outdated systems can request both excessive and irrelevant data, extending integration times for little benefit. Seventy-four percent of 126 FIs surveyed in Bangladesh, India and Sri Lanka, for example, reported poor data quality as one of their top five AML compliance challenges.

Another major obstacle to AML is rating clients on the risk of financial crime. FIs should perform Know Your Customer (KYC) checks both during onboarding and continuously throughout the relationship, as analyzing customer transaction patterns can help uncover fraudsters. However, this forces vendors to access and understand a lot of data, and the manual processes and more basic technological tools that many FIs still use are unlikely to keep pace.

Another challenge is that some technologies are simply not precise enough for the job at hand, with 51% of FIs surveyed saying their transaction monitoring tools produced too many false positives. Old-fashioned rules-based systems monitor customer behaviors for certain activities typically correlated with fraud, prompting compliance teams to manually review any red flags. However, this static approach cannot always keep pace with the complexity and innovation of modern fraud systems.

Leveraging Aging Advanced Technologies in Anti-Money Laundering Efforts

Advanced tools that automate verification, completing the process quickly to preserve the customer experience, can be a game-changer for FIs seeking to comply with AML requirements. More and more financial service providers are turning to technologies such as artificial intelligence (AI) and machine learning (ML) to detect and predict fraud accurately, dynamically and transparently. Forty-five percent of FIs surveyed in 2020 planned to implement AI and bots over the next two years to improve detection while reducing false positives, for example.

ML can analyze data faster than human workers, uncovering patterns that might otherwise be missed, whether that’s indicating bad behavior or confirming the legitimacy of honest users. These systems hone their detection capabilities as they examine more data, so they scale and become more accurate over time.

The biometric approach

Improving data analytics to fight AML compliance is only half the battle, and data collection can be streamlined as well. Biometrics can quickly assess unique credentials, but these methods don’t have universal appeal: while about 60 percent of Americans would be willing to provide fingerprint scans for account security to their banks , only 43 percent of Canadians would say the same thing, according to a recent research report.

Additionally, the specific biometric details collected are important, with 68% of UK consumers willing to verify financial accounts using fingerprints, but only 37% and 36% inclined to undergo iris and skin scans. face, respectively. Voice biometrics are attracting the attention of financial institutions such as Korea’s IBK bank, which reportedly used the method to confirm the identity of called customers within 15 seconds. The bank believes the method is particularly useful for verifying older customers who may have difficulty providing fingerprints on smartphones or using other types of authentication. FIs adopting the biometric approach will need to listen to their specific clients and choose their methods accordingly.

Retail banks, lenders and other financial service providers will still need to navigate two key goals: fast and smooth customer access to their services and robust, regulatory-compliant identity verification measures. Innovative technologies with powerful analytics can enable vendors to deliver more seamless security while providing customers with the reassurance of enhanced fraud protection.

The future of identity verification in the financial industry

More consumers are expected to turn to online channels for their banking and payment needs over the next few years, giving FIs a key opportunity to attract new customers. It is important to recognize, however, that many customer verification methods that have long been used by banks are no longer viable as the online banking world grows. A recent study found that 1.1 billion consumers worldwide lack traditional identity documents, including birth certificates and passports, which could prevent them from opening online bank accounts.

Banks must therefore develop verification solutions that do not rely on such data to follow the flow of consumers who want to access financial platforms through digital and mobile channels. Identity verification measures aimed at consumers, such as biometrics, in addition to AI and back-end ML tools, are likely to play a key role in authenticating these consumers – and banks seem to be doing it. to acknowledge.

Biometrics are also becoming an integral part of identification solutions for businesses outside of banking, as the technology could potentially be used in government efforts to create so-called national identities to protect against digital fraud. These national identities are unique electronic identifiers that can essentially replace the use of identifying information such as driver’s licenses or online social security numbers. About 70% of executives polled in a recent survey believed that having a digital national identity program would give low-income users more opportunities to participate in online banking. These credentials can also be combined with biometric identifiers to create much stronger online profiles of individual users, helping businesses distinguish them from fraudsters.

FIs can also leverage automated technologies to help strengthen their identification efforts, leveraging AI and ML tools to automatically analyze identity metrics and other data for faster results and clearer. Seventy-seven percent of financial executives predict that AI will be the most important technology for the banking space in the next few years, for example, and 66% of banking executives believe that ML tools, solutions blockchain and the Internet of Things will play an important role. role in space.

It is important for FIs to examine how these emerging technologies can help them simultaneously enable seamless onboarding and login experiences for legitimate users while preventing fraudsters from exploiting high-tech solutions. Staying on top of the changing developments in online identification will prove essential in helping FIs compete in digital banking, both now and in the future.


Source link