Sign up for our newsletter and get the latest big data news and analysis.

AI Usage in Banking is Forcing the Conversation around the Ethical Use of Data

Over the past few years, the financial services industry has made huge strides in adopting new technologies like artificial intelligence into its workflow. The cache of data that banks can access holds endless potential to create valuable new products and experiences for clients. With this progress, however, comes concerns around the ethical use of personal data in banking. But with the right business practices in place, banks can reap the benefits of AI while keeping customers in control of their data and protected from its misuse.

Using data for good

According to a recent report from Business Insider Intelligence, banks could see an estimated $447 billion in cost savings by 2023 from AI applications. Some retail banks have already implemented AI features including 24/7 chatbots, virtual voice assistants, predictive analytics and fraud-flagging services.

Business banks are also using AI technology to detect and prevent payments fraud, aid in underwriting assessments and improve anti-money laundering (AML) and know-your-customer (KYC) regulatory checks. The U.N. estimates that around $2 trillion is laundered globally each year, or about 2-5% of global GDP. As money launderers became more sophisticated over the years, AML compliance-related costs rose by more than 50% from 2015-2018. New AI platforms are helping financial institutions stop money launderers, while saving them time and money in the long run.

Driving new efficiencies with ML

Similarly, machine learning (ML) algorithms assess past behavior in order to identify potential trends and future outcomes, as well as flag unusual or suspicious activity. ML tools can also analyze and verify potential customers to ensure they meet strict KYC requirements.

JPMorgan Chase’s contract intelligence platform COiN can review and extract relevant data from about 12,000 credit agreements in seconds — a process that typically requires up to 360,000 hours of manual work per year. These efficiencies reduce tedious manual labor, allowing financial services employees to focus on more high-value tasks — like people management, strategic planning and technology testing — resulting in lower operational costs and higher profits.

AI outcomes must be fueled by ethical data

AI algorithms that run on access to consumer data raise all kinds of questions about the sourcing of that data. Intentional practices, like sharing or selling customer data without consent, are the most negligent form of potential misuse.

Some countries around the world are doing work to better establish the data ownership relationship between banks and clients. Australia has legislated a Consumer Data Right (CDR) and Canada recently launched a ten-principle Digital Charter. Closer to home, the Consumer Financial Protection Bureau (CFPB) released guidelines on data sharing and aggregation.      

Customers deserve to have their data protected across all aspects of their lives. All data exchanges between a financial institution and its customers should be permissioned explicitly, transparently and on a granular-use basis. Data exchange arrangements should be easily available for review, revision or revocation by customers. Customers should also be able to easily inspect the past journeys their data has taken, even if such data has been anonymized or aggregated.

Understanding the spectrum

In the end, the ethics of using AI in banking falls along a spectrum. Some practices, like discrimination and data use or data sharing without consent, are clearly a misuse of customer information and an abuse of their trust. Other practices, like tailoring services, fraud detection and predictive analytics, are of benefit to both banks and their customers.

Financial institutions adopting AI should establish a framework that protects their customers and validates the decision-making behind their algorithm applications.

About the Author

Lisa Shields is the Founder and Chief Executive Officer at FISPAN, where she leads the company with a dual emphasis on people and product. Lisa is an engineer by trade and an entrepreneur at heart, having founded and led Hyperwallet for 15 years before launching FISPAN.

Sign up for the free insideBIGDATA newsletter.

Leave a Comment

*

Resource Links: