It may seem like artificial intelligence fatigue is burning out professionals across the banking space, but new data from American Banker finds that executives are as eager as ever to delve deeper into the technology.
The survey polled 694 experts across the accounting,
Seventy percent of banking and payments organizations report actively exploring or using AI tools to some degree. Roughly 9% of banks and payments companies are aggressively rolling out AI campaigns company-wide, 30% are focusing on small-scale implementations for specific use cases and 31% are taking an incremental approach to AI adoption.
Recent examples of AI campaigns include JPMorganChase's
Read more:
Usage has gone up, and so have
These costs can be a greater barrier for growth than many think, as almost half of bank and payments respondents said the financial burden for updating core information technology was the main hurdle to realizing some degree of positive ROI for AI investments.
Vik Sohoni, senior partner and global leader of McKinsey's banking digital analytics practice, told American Banker that AI use cases in identifying fraud and helping call center agents has "certainly [generated] some returns on investments," but the impact of these tools on a bank's bottom line is much harder to visualize.
"Are these [returns] quantifiable, tangible and visible in the bottom line or are they ephemeral and a few minutes here and there? So far it's more the latter, with some exceptions like call centers," Sohoni said.
Read more:
These increases aren't stopping banks and payment companies from doubling down on AI. Roughly 80% said their organizations were increasing tech spend on AI from less than 10% to more than 25% across the next 12 months.
What follows are more insights into how banks are adapting to the next wave of AI tools, and what technology is next on the horizon.
AI adoption has swept across Bank of America
Most of Bank of America's 213,000 employees are using artificial intelligence, be it the bank's virtual assistant
Bank of America leaders said that roughly 90% of the bank's workforce uses Erica, which purportedly has cut down on IT help desk calls by more than half. Generative AI adoption has also grown among those working at Merrill Lynch and BofA's private banking and call center departments.
"When we think about build versus buy, we don't want to be building things that are increasingly foundational and are available as a commodity," Hari Gopalkrishnan, head of consumer, business and wealth management technology, told American Banker. "We want to leverage the heck out of innovation happening in the industry [and] we want to leverage the heck out of partnerships we have that are investing tons of money in the space that said they have to eventually fit in with our framework."
Read more:
Putting AI at the heart of Lloyds Bank's plan for the future
The London-based Lloyds Banking Group is using Google Cloud to develop generative and agentic AI models to help Chief Data and Analytics Officer Ranil Boteju's plan to "enable the whole bank with AI."
Lloyds' on-premise, legacy machine learning and data science stack was roughly 10 years old when executives began working with Google's Vertex AI platform to build new models a few years ago. Since then, the bank has successfully transferred 15 modeling programs to the Vertex AI platform hosted on Google Cloud, allegedly shaving 27 tons off the estimated lifetime carbon footprint of the machine learning platform.
Being able to use a variety of models through Vertex was a selling point for Boteju when planning the bank's foray into generative AI.
"We can use Gemini when we want, but also the fact that we can bring in open source models, or any number of different models, that is incredibly useful and powerful for us," he said. "It means that we can select the right element for the right task."
Read more:
IBM debuts quantum encryption, AI accelerators in new mainframe
Last month, IBM unveiled its new z17 mainframe capable of running artificial intelligence models at high speeds while also securing transactions and other data with quantum-grade encryption.
Officials say the IBM z17 is capable of processing roughly 35 billion transactions per day through its Telum II chip, made possible by what the company's higher ups say are AI accelerators and 40% more on-chip capacity when compared to the chip in the prior z16 model.
"We know fraud has so many forms and you need a model that's an expert in each one of those forms, and now you can run them all in a single transaction," Tina Tarquinio, chief product officer for IBM Z, told American Banker. "The system really needs to be able to detect the difference between fraud and a valid transaction, so you spend more time on the things you should and less time on the things you shouldn't."
Read more:
Mastercard wades deeper into agentic AI
Mastercard is upping its agentic AI presence through the launch of Mastercard Agent Pay, which aims to help merchants, consumers and issuers alike address needs from purchase opportunities to supply chain management.
Through the new platform, Mastercard uses its data to help generate a host of purchase ideas for consumers managing events, assist merchants with supply chain duties and guide merchants through developing a marketing or sales framework. Technology partners include Microsoft for scaling, IBM for B2B technology and both PayPal and Checkout.com for security.
"Mastercard is transforming the way the world pays for the better by anticipating consumer needs on the horizon," Jorn Lambert, chief product officer at Mastercard, said in a
Read more:
IBM enlists Box to further enterprise AI adoption
Box, a cloud storage company, is working with IBM to strengthen its push into the enterprise AI markets.
The effort combines Box AI, an agentic AI tool that allows businesses to develop and tailor their own closed-system AI models, with IBM's watsonx.ai. Box AI users now have access to a wider variety of models through watsonx, and IBM employees get access to Box AI tools through watsonx in exchange."Financial services firms can accelerate fraud detection and risk analysis by extracting patterns from large volumes of transaction data, enabling real-time anomaly detection and reducing remediation time, all while ensuring compliance and auditability," Dawn Lauter, senior team manager of product marketing at Box, said in a
Read more:
