Artificial intelligence (AI) in banking is not new. Banks are already using AI in heavilymanual processes for accuracy, efficiency, speed and cost benefits. What is new, however, is the move of AI beyond process to interaction. By automating the interaction between all of the different tools and people in an IT environment, for example, a New York-based investment bank achieved a 93 percent reduction in average resolution and fix time (from 47 minutes to 4 minutes).15 Thirty percent of bankers are currently using intelligent virtual assistants on large scale projects or broadly across the organization. The next stage of AI in banking will be toward simpler, smarter interfaces: machine learning to reengineer back-office processes, and AI tools that allow better interaction with customers.
Banks can use advanced machine-learning— systems that gain knowledge from data as “experience” and adapt to apply what is learned in upcoming situations—to streamline back-end processes and support networks. For example, text-mining algorithms pointed at data from internal text, voice, social media and other data sources can help identify unusual and potentially risky or illicit trading patterns. Machines can learn from that data analysis to help banks better detect fraud. A prominent global bank intent on seizing first-mover advantage in its digital strategy successfully tested the IPsoft Amelia AI platform’s ability to support its network of mortgage brokers by providing guidance on policy details.
Tapping AI-enabled tools (like centralized platforms/assistants or messaging bots) to enhance front-end services is a gamechanger. For example, Capital One® Bank developed a “skill” for the Amazon Echo’s® Alexa, allowing people to check their accounts and pay credit card bills via the Echo device.17 Customers of HSBC® can connect with the bank’s virtual assistant, Olivia, to get answers about their credit cards or current accounts.18 “Collette” is one of Accenture’s premier virtual mortgage advisers that uses cognitive science, AI and user-centered design to provide tailored advice on complex mortgage applications— articulating its own thoughts, understanding the intentions and emotions of the customer, making recommendations, and, if necessary, referring the customer to a human assistant.
Most banks get it: 79 percent agree that AI will revolutionize the way they gain information from and interact with customers; 29 percent believe it is extremely important to offer their products/services through centralized platforms/assistants or messaging bots; 76 percent believe that in the next three years, the majority of organizations in the banking industry will deploy AI interfaces as their primary point for interacting with customers; and 71 percent believe that AI is capable of becoming the face of their organization or brand. Basically, AI is creating a new era of computing, rapidly moving from mobile-first to AI-first in the customer experience and moving staff to more judgment-based and higher value added roles.
For both front- and back-office applications, banks are focusing on several AI-related technologies to progress forward. They expect to invest in the following capabilities extensively over the next three years:
• Embedded AI solutions (40%).
• Computer Vision (40%).
• Machine Learning (38%).
• Natural Language Processing (37%).
• Robotic Process Automation (34%).
Bankers do expect high returns on such investments. Their top three reasons for embedding AI into user interfaces: data analysis and insight (60 percent), productivity (59 percent) and cost benefits/ savings (54 percent). Interestingly, 67 percent of US bankers place cost benefits/savings at the top of their reasons to invest in AI. Bankers also expect AI to accelerate technology adoption throughout their organization (80 percent) and ease use of and simplicity in the user interface to ensure a more humanlike experience (78 percent). Seventy-six percent believe organizations will increasingly compete on the ability to make technology fade, or appear invisible to the customer. Reaping such benefits will require banks to tackle privacy issues in embedding AI into user interfaces (38 percent), integration/ capability issues between AI and current IT (36 percent), and data quality (36 percent).