It’s BAI’s 2018 Trend of the Year and is all over the agenda at the upcoming BAI Beacon financial services conference because it underpins digital transformation initiatives. Banks worldwide now use internal and external data sources to develop AI algorithms that improve customer experience, increase agility, reduce operating costs, improve risk management and reduce fraud and financial crimes.
Banks of almost all sizes, especially large global and regional banks, have AI projects underway. These projects should benefit the banks’ bottom lines. For instance, AI is helping banks understand the needs of each digital customer in a way that mimics conversations in branches. Banks that communicate their understanding of customer needs through positive interactions can expand customer relationships, maintain trust and ultimately increase revenue.
Banks are also using artificial intelligence to grow lending usage and improve credit decisions via alternative external data (“alt-data”) as well as credit scores. As they do, they identify ways AI can improve business processes and thus significantly reduce operational expense, even as they speed cycle times and reduce error rates.
But with this fast-paced AI adoption, have banks stopped to ask what their consumers think about it?
Consumers: From health care to financially aware
In a recent survey, SAS found that 69 percent of U.S. consumers are uncomfortable with banks using AI to access credit history for credit card recommendations. And 66 percent are uncomfortable with banks leveraging AI to provide financial guidance. Sixty-five percent weren’t confident that their personal data used for AI is securely stored. The recent Facebook/Cambridge Analytica data misuse has likely only heightened those concerns.
Consumers’ uneasiness regarding AI in banking stands in sharp contrast to health care, where 60 percent of consumers are comfortable with doctors suggesting treatment through AI analysis of their medical information. So while consumers don’t want AI to provide financial guidance, 61 percent are willing to receive lifestyle assessments and recommendations from health care providers based on wearable device data. These positive health care survey results point to the top consumer concern about AI: lack of human interaction when advice or recommendations are delivered. In the health care examples, AI analyzes data, but a care provider discusses recommendations with the patient.
AI and handing off the ball to humans
In light of these survey findings, banks should design AI into customer interaction and service applications in a way that delivers self-service and efficiency for the customer’s benefit—while providing a simple, fast and friction-free transition to human interaction. Designers of customer service chatbots and robo-advisors shouldn’t have to relearn the lessons of customers getting lost in automated support telephone trees, pressing “0 “for a human and waiting an eternity while satisfaction plummets.
A second banking lesson is that AI requires oversight and governance, just like any other form of advanced analytics that touches customers. U.S.-based banks are already familiar with the Federal Reserve’s Guidance on Model Risk Management (SR 11-7), which describes the processes required to govern and manage the use of models and their inherent risks. The use of model risk management, with a centralized model inventory and model assessment capabilities, should inform the bank’s external communications on AI governance, which can address customer fears surrounding personal data protection and ethical use of AI algorithms.
In the survey, banking did receive positive marks for AI use in one high concern area for customers: 61 percent reported they’re comfortable with AI monitoring their online financial behavior to identify fraud and potential threats. This result points to two positives for banks. First, it shows that customers continue to view banks positively. Second, customers are willing to give up some level of data privacy when they see a clear personal benefit.
Putting it all together: Human interaction, consumer satisfaction
Taken together, these survey results make it clear that while banking customers will accept AI automation that protects them from fraud, they want human interaction for advice on their finances.
What does this mean for banks? The expanding use of AI for process automation and decision making—for fraud and similar areas—should continue as it improves the bank’s image as a trusted financial intermediary.
Consumers don’t want to see the financial equivalent of Terminator 3:Rise of the Machines. They want human interactions. Banks already have skilled employees who can deliver excellent advice and customer service, while also boosting productivity with AI.
While AI will continue to revolutionize the banking industry, consumers have made it clear that they want the human touch. Understanding the needs of every customer (and customer experience) should remain priority one for banks—even as you aspire to put your customers first.
While financial services firms may lag other industries in AI deployment, they are better than average at data collection and verification. In this report, learn how the industry has potential to lead by further investing in AI infrastructure and computing...
Persistent inflation and higher interest rates will challenge banks’ ability to meet capital needs and cash flow. That means treasury departments need digital solutions that are timely, capture data from across the institution and anticipate changing economic trends.