Across the global banking sector, a quiet reckoning is underway: the promise of artificial intelligence has outpaced its proof. A new report finds that while financial institutions are racing to invest — with AI spending projected to grow tenfold to $368 billion by 2032 — only one in ten banks is translating that commitment into meaningful, enterprise-wide returns. The institutions breaking through are not those with the most experiments, but those disciplined enough to bind every deployment to a measurable outcome, treating accountability not as a feature but as a foundation.
Only 10% of Banks Monetizing AI Despite Massive Investment Surge
The issue isn't experiments. It's accountability.
Mark
So the headline is that only 10% of banks are actually making money from AI. But the report also says 77% of executives see positive ROI in the first year. How do those two things fit together?
Mimi
The first-year ROI is real—pilots often work. But that's not the same as enterprise-wide impact. Most banks run AI experiments that show promise, then struggle to scale them across the whole organization. The 10% figure is about institutions that have moved past that bottleneck into actual production-scale revenue generation.
Luke
Right, but I want to be careful here. The 77% figure comes from what executives report—which is their perception. The 10% figure is about "significant, measurable ROI," which is a higher bar. We don't know if the 77% includes banks that saw small gains in pilots but never scaled them. Those could be the same population.
Mimi
That's fair. The report is essentially saying that perception and reality have diverged. Executives think they're progressing, but the measurable outcomes tell a different story.
Mark
What's actually different about the banks that are succeeding? Is it technology, or is it how they're organized?
Mimi
It's organizational. The successful banks anchor AI to specific revenue outcomes from the beginning. They build accountability into the system. And they're willing to partner with external providers on a shared-outcome basis—paying for results, not tools.
Luke
That's a structural shift, not a technical one. But I'd want to know: are those banks succeeding because they have better AI, or because they have better governance and accountability structures? The report seems to suggest it's the latter.
Mimi
The report does suggest that. It says the issue isn't experiments—it's accountability. The banks that are winning are the ones that treat AI as a business problem, not a technology problem.
Mark
DBS Singapore generated $565 million from AI in 2024. Is that typical for a leading bank, or is that exceptional?
Mimi
That's exceptional. DBS is one of the most advanced banks in the region. But the point is that it's possible—and it's happening now, not in some distant future.
Luke
Though we should note: we don't know what portion of DBS's overall revenue that represents, or how much they spent to generate it. $565 million sounds large, but context matters.
Mark
What happens to the banks that don't make this shift? Do they fall behind?
Mimi
Almost certainly. If AI spending is going to grow tenfold over the next decade, the banks that can't translate that into revenue will be wasting enormous capital. The competitive pressure will be intense.
The Pulse
- A striking disconnect has emerged: three-quarters of banking executives claim early positive returns from AI, yet only 10% of institutions are achieving real, scalable impact — suggesting widespread self-deception or measurement failure at the leadership level.
- Most banks remain trapped in an endless pilot phase, running AI experiments that never graduate into revenue-generating operations, burning investment without building competitive advantage.
- The banks pulling ahead — in Singapore, the Gulf states, and Latin America — share one discipline: they anchor AI to specific revenue targets from day one, build shared-outcome partnerships with providers, and assign clear ownership over results.
- DBS Singapore's $565 million in AI-generated revenue across 350 use cases in 2024 signals what production-scale execution actually looks like, setting a benchmark that rivals across Southeast Asia, the Middle East, and Latin America are scrambling to match.
- With BFSI AI spending on a trajectory toward $368 billion annually by 2032, the window for leisurely experimentation is closing — execution speed is becoming the defining competitive variable.
Across the global banking sector, a quiet reckoning is underway: the promise of artificial intelligence has outpaced its proof. A new report finds that while financial institutions are racing to invest — with AI spending projected to grow tenfold to $368 billion by 2032 — only one in ten banks is translating that commitment into meaningful, enterprise-wide returns. The institutions breaking through are not those with the most experiments, but those disciplined enough to bind every deployment to a measurable outcome, treating accountability not as a feature but as a foundation.
A new report from Dyna.Ai, developed with GXS Partners and Smartkarma, has exposed a sobering gap at the heart of banking's AI moment. Global spending on AI across banking, financial services, and insurance is on course to grow tenfold — from $35 billion in 2023 to $368 billion by 2032 — yet only one in ten financial institutions using advanced AI is achieving significant, measurable returns. Three-quarters of executives report early positive results, but that optimism rarely scales into enterprise-wide impact, revealing a gap between perception and reality wider than most anticipated.
The banks that are breaking through share a defining discipline: they treat AI not as a technology initiative but as a revenue commitment. They tie every deployment to a specific business outcome, build accountability into their structures, and design partnerships where providers are paid for results achieved rather than tools delivered. This shift — from experimentation to what the report calls production-scale execution — is what separates the leaders from the majority still cycling through pilots.
The evidence is regional and concrete. In Southeast Asia, DBS Singapore generated $565 million in revenue from 350 AI use cases in 2024, targeting $745 million by 2025, powered by a young mobile-first population and a $300 billion SME financing gap. In the Middle East, sovereign ambition and fintech momentum are driving AI into wealth management and cross-border payments, with PwC estimating AI could add $320 billion to the region's economy by 2030. In Latin America, where over 200 million adults remain outside the formal financial system, banks like BBVA Mexico are using AI-driven credit decisioning to extend access to previously excluded populations while managing fraud risk.
Across all three regions, the obstacles are consistent — fragmented data, unclear governance, and cultural resistance — but the institutions succeeding reframe these as accountability problems rather than technical ones. They embed governance from the start and resist the comfort of endless experimentation. As the industry accelerates toward a $368 billion annual commitment, the competitive advantage will belong not to the banks with the most pilots, but to those that can move fastest from proof of concept to revenue.
A new report on artificial intelligence in banking reveals a striking gap between investment and results. While banks worldwide are pouring money into AI systems at an accelerating pace, only one in ten financial institutions using advanced AI are actually seeing significant, measurable returns on that spending. The finding comes from research released by Dyna.Ai, developed alongside GXS Partners and Smartkarma, and it cuts against the optimism that typically surrounds AI adoption in the sector.
The numbers tell the story. Global spending on AI in banking, financial services, and insurance is projected to grow tenfold over the next decade—from $35 billion in 2023 to $368 billion by 2032. Yet despite this massive acceleration, most banks remain stuck in the pilot phase, running experiments without translating them into revenue-generating operations. Three-quarters of financial services executives report positive returns within the first year, but that early success rarely scales into enterprise-wide impact. The gap between what executives believe is happening and what the data actually shows is wider than most anticipated.
The banks that are breaking through share a common approach: they anchor AI deployment to specific revenue outcomes from the start, rather than treating it as a technology problem to be solved. They build accountability into the system, making sure someone owns the results. And they structure partnerships where success is shared—paying providers based on outcomes achieved, not tools deployed. This shift from "Results-as-a-Service" thinking changes how organizations approach the move from experimentation to production-scale execution. When a bank's AI system is responsible for generating measurable revenue, the entire operation tightens.
In Southeast Asia, where mobile banking is ubiquitous and regulatory frameworks are supportive, banks are already demonstrating what this looks like at scale. DBS Singapore generated $565 million in revenue from 350 AI use cases in 2024 and is targeting $745 million by 2025. The region's young, mobile-first population and the $300 billion gap in financing for small and medium enterprises create both the demand and the opportunity for AI-driven solutions. Banks across the region are deploying AI through mobile and digital channels, where customer interaction happens naturally.
The Middle East is pursuing a different path, driven by sovereign ambition and fintech momentum. PwC estimates that AI could add $320 billion to the region's economy by 2030, with financial services playing a central role. Early wins are emerging in wealth management and cross-border payments, where AI is being used to manage client relationships at scale, strengthen compliance, and speed up regional transactions that have historically been slow and unreliable.
Latin America faces a distinct challenge: financial exclusion and fraud risk. Over 200 million adults in the region remain outside the formal financial system. Banks like BBVA Mexico are using AI-driven credit decisioning and fraud prevention to extend lending access while maintaining risk discipline—essentially using AI to include people who were previously considered too risky to serve. The technology becomes a tool for both growth and social impact.
Across all three regions, the institutions moving fastest to production are confronting the same operational obstacles: data fragmentation, unclear governance, and resistance to adoption. But the ones succeeding treat these not as technical problems but as accountability problems. They embed governance from day one. They tie every deployment to a measurable business outcome. They move past the comfort of endless pilots into the harder work of production-scale execution. As AI spending accelerates toward $368 billion annually, the competitive advantage will belong not to the banks with the most experiments, but to those that can move fastest from proof of concept to revenue.
Notable Quotes
Most banks believe they are progressing with AI, yet research shows only 10% of the organizations using agentic AI are seeing significant, measurable ROI.— Tomas Skoumal, Chairman and Co-founder of Dyna.Ai
Many executives told us they underestimated the effort required beyond the pilot stage, and were increasingly open to partnering where execution and ownership could be shared.— William Hahn, Director at GXS Partners
Want the full story?
Read the original at The Manila Times ↗
Economy & Finance