• Evident Insights says bank AI papers rose sevenfold since 2019
• Focus shifts to agentic AI and applied use cases
• JPMorgan 37 percent of publications, Capital One 14 percent
• Wells Fargo and RBC 5 percent each, TD 4 percent
• Use cases doubled in H1 2025 versus H2 2024
• Citi expands Stylus Workspaces with agentic tools
• RBC’s Atom model targets value by 2027
• FICO unveils models for reliability and governance
• Banks investing in talent, cloud, cyber and data
• Research now converting to production at scale

Big banks are reaping returns from a burst of artificial intelligence research, with output and deployments climbing in tandem, according to an Evident Insights study published last week tracking 50 of the world’s largest financial institutions.

The report found the volume of bank-authored AI papers grew sevenfold between 2019 and 2024, topping 750 last year. 

At the same time, focus narrowed toward applications such as agentic AI, and away from computer vision, scientific discovery and biomedicine.

“This isn’t research for research’s sake: it’s laying the foundation for faster deployments, smarter trading agents, and the next frontier of AI-driven financial services,” said Alexandra Mousavizadeh, co-founder and co-CEO of Evident. 

“The leading banks are pushing the frontier on emerging technologies like agentic AI – building the architectures and workflows that will soon underpin real-world applications.”

The acceleration follows heavy investment in tech talent and modernised IT to overcome early generative AI hurdles in a highly regulated sector. 

Firms boosted hiring in AI model development, platform engineering and programme delivery, while increasing spend on cloud, cybersecurity and data infrastructure, according to industry surveys.

Evident’s analysis highlighted five banks responsible for roughly two thirds of sector research output in 2024. 

JPMorgan Chase led with 37 percent of publications, followed by Capital One at 14 percent. Wells Fargo and Royal Bank of Canada each held a 5 percent share, with TD at 4 percent.

As papers multiplied, the number of AI use cases introduced by banks tracked by Evident doubled in the first half of this year compared with the second half of 2024, signalling faster movement from lab to production.

Citi this week added agentic tools to its Citi Stylus Workspaces, an in-house research assistant powered by Anthropic’s Claude and Google Gemini. 

The bank gave developers an AI coding assistant in January, and last month its wealth unit equipped staff with two AI-enabled advisory platforms.

Royal Bank of Canada credited its RBC Borealis research institute with securely training a financial services foundation model called Atom, projected to contribute to RBC’s goal of generating at least 700 million dollars in value from AI by 2027, according to a July announcement. 

Separately, FICO announced two AI models designed to address reliability, transparency and governance concerns.

“The focus among the major players is on research that targets banking processes,” Mousavizadeh said.

“Through their research programs, banks like JPMorgan Chase, Capital One, RBC, Wells Fargo and TD Bank are setting the tone for how AI will be deployed in high-stakes, regulated environments,” she added, noting the sector is moving from research pipelines to production at scale within two to three years.