AI is fundamentally reshaping banking: Oliver Wyman insight
While AI adoption is picking up pace in the financial sector, many organizations have not managed to keep up in terms of governance, client strategies, and organizational models. In a comprehensive set of insights, Oliver Wyman explores various dimension of this issue, like AI agents, the importance of trust, and the possibility of the AI bubble bursting.
"The future of financial services will be determined less by technology itself than by the strategic choices leaders make in the next few years," said Dylan Walsh, partner at Oliver Wyman.
"AI has the potential to reset the economics of the industry – reshaping cost structures, redefining customer interaction, and reallocating value across the financial system at unprecedented speed."
AI agents and customer service
Agentic AI is quickly becoming the primary interface for financial services, potentially pushing traditional banks into the background when it comes to customer relationships. This shift is already evident in the financial sector, but also in other sectors like e-commerce: Agentic AI-powered commerce is expected to account for at least 12% of all e-commerce by 2029.
Unlike traditional search, agentic AI acts as an adaptive problem solver that coordinates complex workflows with minimal human input. This creates a new competitive landscape where "Generative Engine Optimization" replaces traditional search tactics to capture customer attention.
Banks must now choose between early direct participation, secondary involvement through digital marketplaces, or doubling down on existing human-centric channels. Ultimately, those that fail to adapt to this agent-led future risk losing significant market share as customer habits keep changing.
Will AI run the bank of the future?
While the concept of an ‘AI-run bank’ suggests complete autonomy, the true challenge for leaders is a total organizational redesign rather than just technical implementation.
Future success requires shifting from static job roles to dynamic, skills-based tasks where humans and AI collaborate seamlessly. This evolution creates a new leadership role – the ‘orchestrator’ – who manages outcomes across hybrid human-AI teams rather than just supervising headcount.
Currently, a significant gap exists between AI investment and workforce readiness, with many institutions deferring critical decisions on governance and accountability. Banks should avoid treating AI as just a productivity tool, instead focusing on rebuilding their operating models around hybrid capacity.
"To succeed, leaders must move beyond a human-versus-technology debate and be explicit about where human judgment is required in each workflow – and who is accountable when humans and AI agents jointly produce outcomes," said Walsh.
Trust as a hidden advantage in banking
When it comes to the AI race, tech companies have the advantage in terms of speed and innovation, but banks may be able to leverage their unique advantage: Trust. A full half of consumers surveyed said they trust banks with safeguarding their data, a higher percentage than in any other industry.
Banks excel in reliability and accountability, operating under strict zero-tolerance standards for errors in core processes like payments and lending. This institutional discipline contrasts with frontier AI models, which can see accuracy drop significantly when tasked with processing long, complex financial reports.
To maintain this edge, financial institutions must encode their specialized know-how into AI journeys that prioritize transparency and human oversight. By proving they can safeguard data and interpret complex intent, banks can turn trust into a competitive edge that tech disruptors cannot easily replicate.
The threat of an AI bubble
The current AI-driven market boom has begun to echo the dot-com era, with the top tech giants accounting for a remarkable 35% of the S&P 500’s total value. That has led to concerns that a crash may be on the horizon. If this equity bubble pops, an estimated $33 trillion in value – more than the total GDP of the United States – could vanish, likely triggering a severe global recession.
An alternative (but equally dangerous) scenario involves the $6 trillion in debt-financed AI infrastructure, which is increasingly held in opaque, off-balance-sheet vehicles. This mirrors the structural risks of the 2008 financial crisis, as banks may be more exposed to this risk than they realize.
In order to evade the potentially catastrophic fallout of these types of scenarios, financial institutions need to have a plan in place. Banks and other institutions should closely study the impact a major drop in equity would have on their businesses. They also need to have a clear picture of their exposure to AI, both direct and indirect.
"Banks and other financial institutions should be conducting rigorous scenario analysis under both the equity and debt scenarios," said Anders Nemeth, managing partner at Oliver Wyman.
"Firms that act early to execute hedges and diversify portfolios will be best positioned to weather the storm. It remains unclear when – or whether – this storm might come to pass. But the longer these equity and debt pressures build, the likelier the chance of a downturn. And this time, given all the warning signs, banks shouldn’t be caught off guard."
