Investment in AI is soaring, yet its real-world utility is still evolving, with many viewing it as an emerging technology.
While the financial sector has been using artificial intelligence in one form or another for several years, the recent uptick in AI-related activity and investment is sharpening the focus on how far and fast these new technologies can be scaled. According to the latest Infosys Bank Tech Index, global banks allocated 29% of their technology budgets to AI in Q3 2024, up from 20% in Q1—an overall increase of nine percentage points.
A study from IDC forecasts global AI investment in systems, services, and platforms to reach $300 billion by 2026, driving a compound annual growth rate of 26.5% since 2022—with financial services anticipated to account for a significant share.
Among the drivers of this surge, arguably, was the 2022 launch of ChatGPT. Since then, according to Goldman Sachs, $45 billion of inflation-adjusted investment has been committed to AI technology in the US alone as of the third quarter of last year.
In this new era, İşbank is at the forefront. “As a pioneer in financial technology, our mission is to deliver seamless, hyper-personalized experiences through the strategic integration of cutting-edge innovations,” says Sezgin Lüle, deputy CEO of the Istanbul-based bank. “Among these, AI stands out as a cornerstone for reshaping the banking sector and redefining customer experience.”
Today, AI is at the core of İşbank’s plans.
“By enabling predictive analytics, hyper-personalized services, and enhanced operational efficiencies,” Lüle says, “AI is not just a technological advancement. It is a driving force for reshaping the financial ecosystem.”
Another institution leading the charge is São Paulo-based Nubank. CTO Vitor Olivier says predictive AI enables it to gain leverage and deliver value in a competitive landscape.
“We felt from the very beginning that it was all going to be about big data,” Olivier says, “about the right infrastructure fetching as much data as possible, applying the right algorithms, the right policies and frameworks to allow us to be more precise at a bigger scale and to deliver higher confidence decisions at a larger scale and a lower cost.”
For the past three years, Nubank has been wielding GenAI tools to interact with customers and help them better understand their financial situation.The neobank expects AI to be a growth driver for both its business and its customers—and not just in its home market. While international growth in the banking sector has largely been through M&A, Nubank is betting it can grow organically across borders through new lower-cost platforms, enabling it to approve more customers, bank more people, and offer more competitive products.
“We were born as mobile native,” says Olivier. “We don’t have any branches, so all our over 100 million customers are banked through the app.”
While the smartphone has put a bank branch in everybody’s pocket, AI puts a banker in everybody’s pocket, providing customized insights and nudging customers to think in ways that generate better decisions.
“I think that’s the next wave,” Olivier predicts. “It’s around optimizing people’s lives through technology and giving them greater confidence that they are making the right decisions to manage their finances.”
Hyperscaling
Nubank has several partnerships in place, primarily focused on operations, productivity, and infrastructure, several of these with hyperscalers: cloud service providers that furnish services such as computing and storage at enterprise scale.
Hyperscalers arguably have made themselves critical to any expansive AI strategy. In the US, they spent around $200 billion on AI in 2024, according to Goldman Sachs, a number it expects to increase to $250 billion this year. For Standard Bank Group, that’s where much of the investment is focused.
“Ultimately, you go from on-premises computing power to third-party hyperscaler computing power and that’s most probably where your investment will be,” argues Standard Bank CIO Jörg Fischer. At this stage, the firm measures its primary AI investment in time rather than money.
As technology advances, Fischer expects it to become an integral part of daily operations. That said, it will be some time before AI’s impact can be claimed to be “profound.” In the meantime, Standard Bank is firmly focused on “next-level” productivity enhancement incorporating AI.
“We are really pushing AI now, and are using it on a daily basis,” Fischer says. That means working with multiple technology vendors. He’s also nervous when it comes to client-facing AI. Human oversight must keep AI from running “totally wildly”—bringing with it a range of reputational risks—from errors, to ethical concerns, or even liability, he adds.
As with previous computing innovations, AI’s benefits depend on confidence levels, making pre-adoption testing essential. Following the “initial exuberance,” says Satish Babu, principal engineer at Standard Bank, banks are addressing the practical question of how to make AI the basis of a robust set of products that address genuine customer needs.
“We do viability assessments early in the cycle, to see if an idea will give us a reasonable return,” he says. “There’s an element of unknown until we do the testing, but we do make quick judgments about return on investment.
“We always look at the hype as ‘the art of the possible’ and then work out how that applies to our situation and if it makes sense for us. There’s definitely an exuberant hype on what the technology can provide, and I believe it will live up to that at some point in time. But we are quite some distance from there.”
For some areas of financial services, the horizon is further off. “Although we expect AI technology will help enhance returns, we don’t see fully automated investment funds in the near future,” Hidekazu Ishida, an adviser to Global Financial City Tokyo (FinCity.Tokyo), says. “It is because good investment judgments are highly subjective and unique, and the current AI technology does not come close.”
That said, some investment managers are trying to utilize AI.
“Just as Japanese chess players train themselves with AI players, fund managers will increasingly use AI technology to gather and process information,” says Ishida. “We hear that some fund managers are using AI to replace sell-side research. We also hear that some are trying to use non-financial data to assess the speed of management change.” Quantitative tools tend to lag behind change in management behavior, but AI, combined with fund manager creativity, will eventually help investors achieve higher returns, he adds.
Uneven Progress
Attitudes toward the promise of AI are far from even across global financial markets. Parts of the sector remain fixated on leveraging AI for incremental productivity gains or competitive advantage, rather than focusing on its potential to disrupt and transform, observes Dennis Flynn, AI strategist and senior research fellow at the Centre for Sustainable Business, University College London.
“By significantly enhancing predictive accuracy,” Flynn contends, “AI could narrow or even eliminate arbitrage opportunities, forcing a reevaluation of the risk-reward dynamics that underpin modern markets. Those who embrace this paradigm shift, rather than clinging to outdated models, will emerge as the real winners. AI should empower us to achieve more with the same resources, not simply do the same with less. Letting go of familiar ways of working is difficult, but we are beginning to see a shift in mindset.”
For many banks, however, AI is already central to improving operational efficiency, enhancing decision-making, and expanding product offerings, with strategic partnerships helping them to scale these advantages and speed innovation.
