For years, the tech industry has dominated the spotlight when it comes to the introduction of brilliant or disruptive solutions to the world stage. We continually expect behemoths like Google and Apple to unfurl the next big trend that will change the way we do business and live our lives. But another industry is starting to have a technological revolution of its own — especially as it pertains to artificial intelligence (AI).
In the banking and financial services sector, AI is cropping up everywhere — in lending, marketing, customer service, fraud detection, anti-money laundering, authentication, cybersecurity, virtual assistance, virtual advice, recruiting, employee training, and more. Recent high-profile examples include the use of Amazon’s digital assistant, Alexa, by UBS for customer service; development of a robot to execute trades for JPMorgan; and the creation of an AI fraud detection team for Morgan Stanley.
The general enthusiasm in the financial industry for AI and machine learning is growing, even if the level of investment and results are mixed at the moment. In April of this year, the Financial Times reported that, “of the seven big banks willing to estimate the long-term cost savings of AI, six said it would cut costs by less than 20 percent,” but pointed out that others were more optimistic.
But even as banks are willing to take a chance on AI by investing in tools like chatbots and virtual everything — spending anywhere from $3 million to an estimated $50 million or more on these efforts — there is another opportunity wherein huge potential lies: data.
Data runs through the bloodstream of every financial institution, making inroads and out roads to and from partners, employees, governments, regulatory bodies, security systems, and especially customers. This flow of valuable information is ripe for new forms of intelligence to parse, predict, and act upon it. In fact, of the top reasons banks are using AI, data analysis and insight come out on top at 60%.
Because data is the real asset of the banking industry, this is where AI can really shine.
Since the customer experience is the ultimate proving ground for banks and financial services, focusing AI development there is one of the more prescient and practical ways to cultivate brand loyalty, grow revenue, and stay competitive.
Here are two customer-focused areas in which banks could consider AI solutions:
Customer profiling. Masses of raw data from each customer are lying in wait to be algorithmically sorted and structured to create hyper-detailed customer profiles. With these profiles in hand, banks can then far more accurately assess risk for each customer, craft more personalized communications, and precisely target each customer with more appealing and appropriate offers and products.
Customer behavior patterns. Coupled with personal data, banks also need to understand how customers are making decisions. Using machine learning to analyze customer behavior and language patterns lets banks predict how, why, and when customers make the decisions and choices they do. Algorithms can then be used to automate those decisions, while predictive analytics can anticipate changes in customer behaviors over time.
Bottom line: using AI to mine customer data can help banks:
Even with big potential for impacting a business, there are certainly some reasonable objections to be had concerning the incorporation of AI into banking.
Banks aren’t wrong to ask if AI is worth the hassle and investment, considering the size of their footprints and the slow-to-adopt nature of the finance world. Banks are massive institutions, after all, with a multitude of teams and departments, hundreds if not thousands of customers, and a dizzying number of security and compliance concerns that cannot be overlooked. So major changes like the introduction of emerging technologies often require an extended period of consideration, eventual executive buy-in, and even more time to successfully — and carefully — implement across the organization.
It’s understandable that banks that are just now entering the AI space are exercising caution. They’re using AI initially for smaller tasks and test projects and making sure that these new projects can actually improve end-to-end processes and customer relationships before they agree to fund full-steam-ahead innovation.
Take UBS, for example. They ran a trial of using chatbots for customer service, and while they saw 60% to 85% “accuracy of understanding,” it wasn’t high enough yet for them to officially adopt the program.
Once larger-scale AI development is greenlighted by banks, it will still require a major effort to deploy enterprise-wide.
That’s why having a robust infrastructure to support high-powered AI applications — one with flexibility and scalability, such as a hybrid solution that combines public and private cloud with on-premises IT — is critical to ensure that deployment goes smoothly and that the right levels of support and expertise are available.