Artificial intelligence (AI) is diffusing into every sphere of humanity, raising concerns about its impact that range from the run of the mill (will it help people live happier lives?) to the dystopian (will it drive humans out of work?). The answers to these questions are a work in progress.
AI is a set of technologies built to mimic and augment human intelligence. In modest terms, it is a system that can recognize the world around it; evaluate and fathom the data it ingests; and take or suggest actions based on that evaluation. Recent developments in data-driven pedagogy and tech have advanced the perceivable application of AI.
From a human-value perspective, AI can be interpreted as a set of complementary tech-driven forces that, when applied correctly, can emulate and interact with humans to augment the value creation lifecycle and solve cognitive problems commonly associated with human intelligence.
Part 1 of this two-part series examines the implications of AI-led disruption. Part 2 will look at ways banks can adopt and leverage AI.
Today’s AI is weak: narrow, highly task-oriented, focused on well-scoped and -defined value components. The often-cited examples of self-driving cars, voice-based AI agents etc., while very impressive when it comes to interacting with humans, are highly restricted by their assumptive models.
For now, AI’s capabilities are limited — there is no “strong AI” that has beaten the Turing test convincingly. The upsurge of AI draws reactions similar to those to the early period of computerization-led Industry 3.0, during which there was a clear shift to computer-led business modeling, greater assembly line automation, and mass production. However, we believe fears that the worker will become an appendage to the machine are unfounded.
The key question is how organizations can harness the potential of AI — i.e., create the next J.A.R.V.I.S to complement human intelligence or fuel the rise of machines that, over time, take on a greater degree of operational responsibility.
By 2020, AI is slated to be a top-five investment priority for more than 30% of organizations, according to Gartner. With Industry 4.0 taking hold in the manufacturing space and greater Internet of Things (IoT) device adoption blossoming across every industry, the digitization value stream has triggered an unparalleled intelligence explosion. The time to reshape the organization’s approach to AI is front and center. In particular, businesses should seize the opportunity to apply machine learning (ML), image analytics, deep learning, and natural-language processing to harness the data spawned by every customer touchpoint, device, social media interaction, and business transaction.
The grand convergence of digitization, AI-led data science, and the underlying computational power that fuels today’s business has set the stage for a wave of creative disruption. This, in conjunction with intelligent automation and pervasive sensing, is upending the experience delivered to customers, business partners and employees. The cumulative influence of these forces is ushering businesses toward new opportunities and reimagining business models centered on AI as a fundamental aspect of value delivery.
Although AI is pervasive across the banking value chain, to grasp its totality we propose these key lenses (see Figure 1) to understand the effect of AI on the banking ecosystem.
Robust decision-making is integral to the banking value chain, covering core segments such as loan decisions and credit modeling through fraud detection. The greater quantity of data is being used to train AI algorithms to support — and in some cases to make — more informed decisions. However, in most cases, the colossal amount of data generated by businesses, combined with processing turnaround requirements, have outstripped the feasibility of human-only analysis. As a result, the emergence of machine and deep learning are poised to improve the quality of decisions and enable real-time decision making.
For example, Goldman Sachs’ investment in Kensho Technologies and the eventual acquisition of the ML startup by S&P Global is an example of how AI can assist human analysts by speeding decision-making in areas such as stock picking, investment analysis, and predicting currency movements.
In 2019, the banking industry is likely to witness around $5.6 billion worth of investment in automated threat intelligence and fraud prevention AI use cases alone.
Globally, banks are under pressure to improve cost ratios. Increasingly, many are either looking at AI-driven solutions to assist human agents in improving throughput, or exploring AI-centered automation to expand the coverage to those processes that require perception-driven decision-making such as clearing and reconciliation of handwritten checks, purchase orders and invoices, etc.
Operational processes, such as collections, are now assisted by ML-based next-best-action analysis to improve performance through a hybrid human-machine workforce. According to FICO Analytic Consulting, a financial analytics company, almost 70% of banks are planning to use AI to improve collections and recovery by 2019.
Banks have adopted conversational AI to drive tailored interactions through chat-bots — digital personal assistants that use natural-language processing technologies to create more interpersonal consumer experiences. Bank of America, for example, has introduced Erica, a voice- and text-empowered bot that helps customers make smarter decisions by suggesting ideas on saving money, reporting credit scores, and encouraging bill payments.
Deploying AI across the banking value chain can free up resources to invest in new ideas and markets, reduce costs, and create new interaction points — thus allowing banks to create new business models and expand into new territories and customer segments that were previously too expensive to serve.
This AI-led business value is demonstrated in the small-denomination-lending and alternative-credit-channels business segments, wherein personalized low-cost robo-advisory services are able to reach much larger segments of non-high-net-worth investors. This has tipped the scales and changed the wealth management industry. It required AI-based customer risk and portfolio profiling, as well as the automation of trade processing, clearing and settlement mechanisms, to get there.
Ardhendu Acharya, Senior Manager, Kumar Ray, Manager Consulting, Amit Anand, Assistant VP, and Madhu Ponnuveetil, Senior Director, at Cognizant’s Banking & Financial Services Consulting practice contributed to this article.