Adoption of AI, especially generative artificial intelligence, will rival or exceed the impact of previous digital revolutions in banking. But the pace of AI-driven change will put those earlier waves to shame. So argues Accenture in its latest report on the future of banking:
“While other tipping points have revealed themselves unhurriedly, gradually winning over the skeptics, the adoption of generative AI is happening with almost frenetic haste.”
In Accenture’s annual compendium of the top 10 trends for banking in 2024, every one in some way relates to adoption of GenAI in the industry. In the words of the report, “these technologies are unlikely to change what banking does, but they will dramatically transform how it does it.”
Each trend, Accenture says, will be “either caused or amplified by AI.”
The GenAI revolution however, may look different in banking than other industries. While digitalization turned retailing on its head — Amazon versus Sears Roebuck, for example — there hasn’t been a fintech or challenger bank that has yet reached the same stature as major traditional banking institutions, according to Michael Abbott, senior managing director and global banking lead at Accenture. And he doesn’t think GenAI will unseat the incumbents this time around either.
A key point will be the tension between automation of banking under GenAI versus the augmentation of human effort by the technology. In an interview with The Financial Brand, Abbott says he believes that humans are going to continue to be involved in many processes, but differently.
Automation Versus Augmentation in the Banking Industry
Accenture examined 19 industries and found that banking ranked highly both among those businesses with higher potential to have aspects automated via GenAI (39% of workload) and among those with higher potential for augmentation (34%). Among the 19 categories, banking ranked least in terms of the portion of industry workload that had low potential for both automation and augmentation (27%).
For the sake of contrast, industries with lower potential for automation and augmentation included aerospace and defense, consumer goods and services, and natural resources.
Much of the controversy about AI in general has been its potential for replacing human workers with software. GenAI has raised the same specter, but with broader and deeper implications given its utility for “brain work.”
Abbott thinks the industry should be looking harder at augmentation possibilities.
“Don’t think of GenAI as replacing bankers all the time,” says Abbott. “Think of it as making them more intelligent. It’s like having Kasparov whispering in your ear as you’re playing chess.”
Read more: How to Mitigate Risk and Consumer Fear As Banks Adopt GenAI
How Augmentation Could Assist Bankers with Lending
Abbott gives a couple of examples of how this could work.
The first concerns a key difference between “traditional” AI and generative AI. Artificial intelligence in the past had to be limited to tasks that involved numbers. Such input is broadly referred to as “structured data.” By contrast, dealing in words, facts, ideas and concepts, generative AI is capable of working with unstructured information, explains Abbott.
Traditional AI would perform loan underwriting by evaluating credit data and other numerically expressed factors with algorithms. Use of machine learning enabled such AI to “learn” and adjust its credit analysis as it did so — automating and speeding an established process.
On the other hand, let’s say a mortgage underwriter had a loan package and wanted to determine if it met all requirements to be able to be put through Fannie Mae or Freddie Mac. “Ordinarily, they’d have to know all the rules and regulations,” says Abbott.
Now, generative artificial intelligence, trained in mortgage lending, can evaluate the contents of the loan package. It wouldn’t handle the underwriting, but would advise the human underwriter regarding what might be missing and if the proposed loan could be brought up to Fannie/Freddie requirements.
“All the things that a loan officer would normally have to search for on their own, generative AI can augment the process and help make it that much faster, easier and better for the lender,” says Abbott. Judgment could still be involved, ultimately, and Abbott says that’s where the banker comes in.
“With generative AI you almost always want to have a human in the loop.”
— Michael Abbott, Accenture
What will be the final blend of traditional AI, generative AI and human remains to be seen. Nonetheless, the firm insists that banking institutions that explore the technology for augmentation will see increases in productivity and revenue.
Read more:
- Lessons from Ally’s Experiment with GenAI Marketing: Commit, But Carefully
- Banks: Be Prepared to Explain ‘Explainability’ of AI
- How Financial Institutions Can Win the Next Stage of GenAI Race
Accenture’s Top 10 Banking Trends for 2024
1. The rise of generative artificial intelligence.
2. Capturing the digital dividend — making more of digital touchpoints.
3. New risks must be on banking’s radar, including the long-term impact of the trends impacting residential and commercial real estate.
4. Evolving ways of working as GenAI embraces rote work and expands beyond that.
5. New abilities to price right down to the individual customer level.
6. Cloud-first thinking will become the strategy for institutions of all sizes.
7. Regulation will evolve as the technology of the regulators and the regulated will be capable of communicating with each other.
8. “Information Technology” as terminology will go on the shelf, replaced by engineering thinking.
9. The boat-anchor of ancient core systems will suddenly not be an excuse for banks anymore, as GenAI will make rapid, bulk conversion of old computer code less daunting.
10. Banking will move beyond traditional Six Sigma process improvement methods, as GenAI will help change facets of banking beyond the numerically measurable.
How GenAI Could Help with Advice, Customer Service and Sales
The banking industry has used “Six Sigma,” an approach to process improvement used to re-engineer banking tasks, for decades. Helpful as it was, its applicability had limits because it focused on numbers — structured data, again
“If you couldn’t measure something, it was difficult to factor it into the system,” Accenture says. “Learning from intuition and experience — the shades of gray — was often asking too much of these methods.”
The report predicts that GenAI will enable companies to go beyond the process improvements Six Sigma made possible.
You Ain't Seen Nothing Yet:
Accenture suggests that the most aggressive banking leaders will use GenAI to reengineer a good deal of how their institutions operate in order to reach the next level of efficiency — basically rethinking the whole bank.
Abbott cites another use case in which GenAI acts as counselor or guru.
When a banker is communicating with a customer or prospect, the human leads the conversation, but the GenAI listens constantly, ready to suggest an alternative approach or even a change of subject if the person isn’t responding to the original pitch.
“It can be like a Google Maps for conversation,” says Abbott.
Consider a relationship wealth manager. Let’s say his or her client clearly has no interest in a home equity loan.
“The GenAI can say, ‘Look, stop taking the conversation down the path of home equity, because it’s not working,'” says Abbott. “‘Maybe you should ask them about their kids and see if college savings is more something you should explore.'”
“Suddenly, the GenAI can take you down a green path,” Abbott says.
Abbott says Accenture’s research suggests that the long-term value of adopting GenAI is not in reducing headcount. “We find that the real value is in taking waste out and putting value in,” he says. The more mundane tasks that technology takes on — “tasks that not even a beginning banker wants to do” — frees them to add value.
Read more: How Community Banks Can Entice and Recruit Software Engineers
When GenAI Takes on the Customer Conversation
Abbott points out that one of the casualties of the digitization is the opportunity to have an actual conversation with customers and prospects.
“Almost 99% of customer touchpoints in banking are now remote ones,” says Abbott. “It’s simply people checking balances, paying bills, etc. They’re not talking to bankers anymore.”
Abbott thinks this situation can be changed in a way that doesn’t take away customer convenience, but regains the conversation. In this case, GenAI and humans team up, but with the technology taking the lead.
One example would run like this: A customer has tried logging into a bank website but has the password wrong. After four or five attempts, the customer tries the bank’s interactive voice response line. Instead of simply handing over the balance, the IVR’s voice might say, “I see you are calling from your phone and that you tried to check your balance via the website. Were you having trouble logging in? I can help you with that if you like.”
“We no longer have to have the functionally correct, but emotionally devoid experiences that we’re pushing customers through today,” says Abbott. The conversation could go further, such as “You have a lot of idle cash in your checking account. I can help you put some of that into a CD. Are you interested?”
Turning “dead-end touchpoints” into potential cross sales would be a major advantage of GenAI. “I believe there’s a massive opportunity for banks to capture what I call the ‘digital dividend,'” says Abbott.
Initially, Abbott sees the GenAI pulling in a human if something wasn’t working out. In time, that wouldn’t be necessary.
“As the GenAI gets better and better, you just won’t need to have a live person,” he says, “unless you have a complex problem.”
Overlaying this capability will be GenAI’s ability to generate thousands of potential offers, tailored to each customer’s needs and issue of the moment.
A key support for this strategy is that banking brands remain highly trusted, according to Abbott. It’s critical that in blending GenAI and human interaction with customers that longstanding trust isn’t blown.
Read more: ChatGPT Will Become ‘ChatOMG!’ in 2024, Forrester Predicts
Will GenAI Widen the Gap Between Mega Banks and the Rest?
Examples cited in the report concern larger banks, some of them in other countries. However, Abbott also sees the technology becoming “democratized” — available to pretty much any size bank or credit union.
“You will no longer need six or seven PhDs to use this technology,” says Abbott. “You just need somebody who knows how to string together the capabilities.”
Much of GenAI technology can be accessed through the cloud, he adds, “and that makes it a great leveler for medium- and small-sized banks.”
Want to go deep on AI best practices for banks?
Attend our AI Masterclass — Unlocking the Power of Artificial Intelligence in Banking — at The Financial Brand Forum 2024 on May 20-22 in Las Vegas. Led by Ron Shevlin, chief research officer at Cornerstone, this three-hour workshop will be jam-packed with lessons learned from industry leaders and real-world case studies.
For more information and to register, check out the Forum website.