JPMorgan Chase and Capital One are far more mature in the development and use of artificial intelligence than their peers, according to a
Research and analysis firm Evident assessed the AI talent, innovation, leadership and transparency of 50 banks in North America, Europe and Asia-Pacific. The team relied on public data and documents, including press releases and research papers, and interviews with leaders at the banks.
North American banks took six of the top 10 positions. In addition to the top three, Wells Fargo, Goldman Sachs and Citigroup all made it into the top 10.
"JPMorgan Chase, Capital One and RBC are very much in the lead," said Alexandra Mousavizadeh, CEO and co-founder of Evident.
A look at how the leaders got to where they are is relevant to all the banks that didn't make the list and that risk being left behind as the top banks achieve cost cuts and new sources of revenue through the use of AI.
JPMorgan Chase's ranking at the top "did not happen overnight," said Teresa Heitsenrether, chief data and analytics officer at JPMorgan Chase, at the Evident AI Symposium at the end of November. "It's been many, many years in the making." She gave credit to the bank's CEO, Jamie Dimon, who appointed her to the C-level AI leadership role earlier this year.
"Jamie was early in the game of understanding what the potential transformative power of AI could be, and he very much used this as a commercial imperative for the firm," Heitsenrether said. "He thinks it's critical for the future success of the company."
Capital One's position in the number-two slot proves that it's not strictly the largest banks that do the most and the best with AI, and maturity is not necessarily in proportion to bank size, Mousavizadeh said.
The bank "is very much an AI-first driven culture," Mousavizadeh said. "It is a cloud-native bank. It has very strong AI leadership. It hired Prem Natarajan to be the new head of AI, who brings an immense amount of experience from big tech. And that permeates the bank, his team and the various lines of business." Capital One's programmers are streamlined, they're organized, they're fast in terms of getting models into production," she said.
The banks that dominate the top of the list, across all four pillars, share certain qualities, Mousavizadeh said.
"It requires a group-wide strategy that goes into getting your ducks in order in terms of the AI talent stack, the innovation stack, but also your leadership," she said. "And we see that feed into the transparency of responsible AI."
Innovation
In the innovation category, Evident looks at the quantity and quality of each bank's research papers, patents, ventures, acquisitions and partnerships with tech companies and universities.
"In all of these, JPMorgan is head and shoulders above everyone else," Mousavizadeh said. "They have the strongest research unit. They have the most papers and citations."
JPMorgan Chase is also active in open source communities like GitHub and Kaggle, Mousavizadeh said.
However, Bank of America and Capital One have more patents than JPMorgan Chase.
"Bank of America has followed a very active patent strategy for its innovation stack, but not so much on research," Mousavizadeh said.
Capital One is "quite far ahead and it's continued to double down," Mousavizadeh said, with the right talent, the right vision, the right operating model and the right innovation stack.
Top talent
The banks that have a clear AI strategy that they can communicate and that can move fast on AI adoption attract employees with AI expertise, Mousavizadeh said.
"AI talent does not want to be in a bank that really gets stuck and bogged down with slow decision making and that's unable to address more complex infrastructure questions," she said.
Evident ranked Capital One number one in the talent category.
"Where Capital One does really well is when it comes down to the proportion of AI talent to the number of employees in the bank," Mousavizadeh said. It also has an AI-first culture and it's a cloud-native bank with a strong AI leadership, she said.
A new model rates large banks on their efforts to develop and deploy artificial intelligence technology.
Capital One works with a network of academic institutions to advance research and theory in applied AI and machine learning, including the University of Maryland, New York University, Columbia University, Massachusetts Institute of Technology, University of Virginia and University of Illinois Urbana-Champaign, according to Natarajan.
"We have a particularly strong belief in partnerships between industry, academia and government to ensure diverse perspectives and equities when developing, testing and deploying AI capabilities," Natarajan said at the U.S. Senate AI Forum on November 3.
Capital One's many collaborations with academia mean "they're at a point where they don't necessarily need banking sector expertise, but they're very much focused on specific AI expertise, whether it's from deep learning or neural networks or synthetic data or causal AI," Mousavizadeh said. "They go to academia to pull that in. And they're doubling down on hiring from academia where many of the other banks just sort of hire from each other, or from consultancies."
JPMorgan Chase, which Evident ranked second in the talent category, has some advantages when it comes to attracting tech talent, including its tech budget – $15 billion a year – and the size of its tech staff, which includes 50,000 software engineers.
The New York bank has hired talented people, created connections with academia and created internships and training to acquire employees with AI expertise, Heitsenrether said.
"We have vast amounts of data and we have the ability to invest," she said. "What better place to work than a place where you can perfect your craft and work in an environment where you're going to continue to be on the cutting edge?"
