Financial services companies increased spending on cloud, data and cybersecurity to support AI efforts in recent quarters, in addition to beefing up skills and talent pools. Now, banking executives are on the hunt for tangible results.
Citigroup, Wells Fargo, JPMorgan Chase and Goldman Sachs led an industrywide AI push that has seeped into back-office operations, in dark corners containing legacy systems and as part of client-facing experiences.
Speaking during third-quarter earnings calls this week, executives from top banks highlighted use cases in pursuit of efficiency.
“We are committed to embedding AI into how we work,” Citigroup CEO Jane Fraser said during the company’s Q3 2025 earnings call Tuesday. “These tools save hours each day by automating routine work, analyzing data and creating materials in minutes instead of hours.”
Citi has capitalized on momentum gained from modernizing underlying technology systems, propelling the business toward AI. Citigroup’s venture capital arm is the most active in AI investments, of the 50 banks tracked by Evident Insights.
Amid deployment efforts, the company strengthened governance last month by adding a new head of AI to partner with its broader executive management team. Nearly 180,000 employees across 83 countries are using the company’s proprietary AI tools, according to Fraser, speaking during the call earlier this week.
Teams throughout the business are using the technology to resolve client inquiries, gain real-time insights, automate code reviews and improve productivity. There have been more than 1 million AI-driven automated code reviews this year, creating 100,000 hours of weekly capacity for developers, according to Fraser.
Other AI use cases include risk reduction and customer experience improvements. Fraser said the company is now “in line with peers” in terms of the level of automation and preventative controls on the back end.
Citi is also piloting agentic AI for about 5,000 colleagues, an effort that launched last month.
“It allows complex multistep tasks to be completed with a single prompt, and the early results are very promising and we’ll expand access to this in the months ahead,” Fraser said.
Efficiency, efficiency, efficiency
Whether they peppered AI throughout the earnings report or only briefly mentioned it, banking executives were clear about their goal to bolster operations with the technology.
Wells Fargo executives only mentioned artificial intelligence once during its latest earnings call, highlighting a single use case: AI as part of automation efforts to drive efficiency.
“When you think about the efficiency agenda … we still think there’s a significant amount to do across the company,” CFO Michael Santomassimo said during the Q3 2025 call earlier this week. “Some of that is people-related, headcount-related, and you can see our headcount gradually coming down quarter after quarter after quarter. We still have more to do there as you start to continue to automate more processes. AI helps on some of that for sure.”
JPMorgan Chase is thinking about AI’s impact on efficiency and subsequent effects on headcount, too.
“For us, running a company of this type, we need to make sure we stay anchored in facts and reality and tangible outcomes,” EVP and CFO Jeremy Barnum said during the company’s Q3 2025 earnings call Tuesday. The company previously estimated they’d have $2 billion in AI-related upside.
JPMorgan commands the biggest AI talent pool of the largest global financial firms, according to an Evident Insights’ annual report. The company has directed significant investment and energy into AI adoption, even before the current generative AI boom, according to Barnum.
JPMorgan’s goal isn’t to make its leaders prove cost savings or productivity boosts from AI, which can push colleagues to “use AI in ways that are actually not efficient and that distract you from doing the underlying process reengineering that you need to do,” Barnum said. Instead, the company will continue to restrict headcount growth.
“We’re going to do the same this year: have a very strong bias against having the reflective response to any given need to hire more people and feeling a little bit more confident in our ability to put that pressure on the organization because we know that, even if we can’t always measure it that precisely, there are definitely productivity tailwinds from AI,” Barnum said.
Efficiencies are also top of mind for Goldman Sachs. The financial firm unveiled a centralized operating model, called One Goldman Sachs 3.0, on Tuesday, that puts focus on AI-driven efficiency.
The multiyear effort will build over time and progress will be measured across several goals, including capacity to scale, risk management and profitability.
“To start, we are drilling in on a handful of front-to-back work streams that can significantly benefit from AI-driven process reengineering and will help inform our longer-term approach,” CEO David Solomon said during the firm’s Q3 2025 earnings call earlier this week.
Responsible adoption
As banking institutions continue to fine-tune their technology strategies, responsible AI practices are critical to success.
More than half of technology leaders at financial services firms point to responsible AI standards as a leading driver of ROI, according to a FICO survey conducted by Corinium and published last month.
“Most organizations are not in a great position today,” FICO Chief Analytics Officer Scott Zoldi told CIO Dive, referring to the current state of responsible AI adoption in banks. “It’s getting better, but it’s nowhere close to where it needs to be.”
Fewer than 13% of organizations surveyed said they had fully integrated key AI development and deployment standards, such as model monitoring and bias mitigation.
Responsible practices go beyond risk mitigation, Zoldi said. A robust governance program will put focus on explainability, transparency, training data and enforcement.
“For FICO, it’s on an AI blockchain so that we can drive adherence to the standard,” Zoldi said. “It’s not just a motto, but really a set of standards that must be met and provide the ongoing rules around how you execute and monitor that AI moving forward.”
Creating a standard isn’t enough. Organizations must ensure standards are communicated and understood by impacted teams.
“Visibility and transparency really empower everyone to understand what we’re trying to achieve with AI,” Zoldi said.
