Artificial intelligence has been among the hottest topics in banking in 2017.
Nearly all of the biggest financial institutions have made a major announcement around new AI applications this year. This is not all hype—few industries are as vulnerable to disruption by AI as banking.
That’s because banking is a process- and data-driven industry. AI tools are ideal for leveraging data to optimize processes. That, in turn, can enable a myriad of benefits for banks, including personalizing the customer experience, automating operations, and improving risk management.
However, overall adoption of AI is still relatively low throughout the industry, with few mature case studies to help guide banking executives in deciding how to leverage AI’s disruptive potential.
Less than a third of traditional financial institutions have initiated an AI project, according to a survey of bank executives released earlier this year by Narrative Science and the National Business Research Institute.
Additionally, the difficulty in acquiring scarce and incredibly expensive AI talent means that many banks lack familiarity with AI. A report by Paysa released in November indicates that financial services companies allocated over $82 million for AI development and research this year, but almost half of that was invested by three very large banks—JPMorgan Chase, Capital One, and Wells Fargo.
Beginning at back-end makes more sense
To get started on the road to leveraging AI, banks need a strategy that addresses the broad variety of AI use cases in the industry, the many challenges to implementing AI, and each bank’s own digital strategy.
For most banks, leveraging AI will start with back-end applications, as the technologies behind those applications are more mature.
Many in banking—and other industries—are fascinated by customer-facing AI applications based on computer vision and natural language processing, like chatbots.
However, it’s the rapid advancements in different types of machine learning technologies—which are well suited to gleaning insights from the mountains of data that banks compile—that are driving innovations in those customer-facing applications.
Additionally, back-end AI applications tend to involve less risk than front-end ones that directly affect the customer experience.
Still, banks have an incredible plethora of use cases to consider in regards to machine learning tools. Prioritizing those use cases will come down to two important factors: the banks’ business priorities, and the data that’s available to feed algorithms.
Where does it hurt?
Every bank has different pain points and business goals that AI can help resolve.
For example, an institution with many high priority commercial clients may explore using AI to help automate receivables, as Bank of America Merrill Lynch is doing in partnership with HighRadius in a program called Intelligent Receivables.
Alternatively, a bank looking to cut compliance costs may use machine learning to automate the collection and analysis of data for money laundering detection. HSBC, for example, has partnered with startup Aysadi, to that end.
The options appear to be nearly endless—AI can help banks more accurately determine credit risk to help grow their lending business; develop more sophisticated investing models to boost their wealth management units; and help small business clients automate expense reporting and fraud detection.
A decade from now, many banks will likely be using machine learning across all of these applications, but mapping out an AI strategy requires prioritizing AI solutions that can help with your bank’s biggest and most pressing problems.
Recipe begins with raw data
However, identifying business needs is only one part of the puzzle to leveraging AI—if the right data isn’t available, machine learning and deep learning tools can’t deliver meaningful insights.
For all their potential, AI technologies are only as useful as the data they are trained with. Banks have tons of data, but that doesn’t mean that the right data is always easily accessible.
Many banks still collect and store enormous volumes of data manually. Even data that has been captured electronically is often siloed in legacy systems, making it difficult to extract and analyze. More and more financial institutions are undertaking the tall task of opening up their massive data sets through cloud migrations, systems upgrades, APIs, and middleware as part of omnichannel and customer centricity initiatives.
Those banks that are further along on this path are better positioned to leverage AI, as these initiatives lay the foundation for gaining new insights and operational efficiencies through machine learning.
In this sense, a banks’ AI strategy is an extension of its digital transformation strategy. Many AI projects will typically start with a long and labor-intensive period of finding, extracting, and cleaning up the data sets necessary for algorithms to draw meaningful conclusions. Those efforts can be drastically simplified if the necessary data is already cleaned up and easy to extract, giving staff more time to focus on fine-tuning analysis rather than examining massive data files.
Efforts must be unified
That’s why banks need to lay the groundwork for leveraging AI by aligning business, digital transformation, and AI strategies. Once business needs are properly identified and prioritized, and digital transformation initiatives are implemented to support those needs by making data more accessible, banks will be well prepared to use AI technologies to tackle their most pressing business problems.