.jpg)
Financial services firms are investing heavily in AI, yet many are struggling to see measurable business results.
A recent PwC survey of more than 4,400 CEOs found that 56% saw no measurable increase in revenue or reduction in costs from AI over the past year.
Getting AI into the business is easy compared to proving it’s worth the investment.
Where the 56% went wrong
The natural reaction is to assume the problem lies with the technology that the models underperformed or that vendors oversold what AI could deliver. However, the evidence points elsewhere. The companies seeing meaningful results from AI are not necessarily using different technology; they are approaching it differently.
Across financial services, many organizations have invested in AI pilots, centre of excellence, and dedicated leadership roles, yet most of these efforts remain focused on making existing processes faster or cheaper. While these initiatives can improve efficiency, they rarely change the underlying business model or create new sources of value.
This is what we call the Modernization Trap: using AI to optimize the business you already have rather than rethinking the business you could become. While many firms continue to focus on incremental improvements, companies like Block are using the data generated through Cash App and Square to build entirely new lending capabilities, demonstrating how AI can reshape a business rather than simply streamline it.
The cost of doing nothing
Treating a structural shift as an efficiency exercise comes with a cost that often stays hidden for years. In businesses such as loan servicing, claims, underwriting, and fund administration, competitive advantage traditionally came from scale, regulatory expertise, and years of operational experience. AI is changing that. As models and workflows take over tasks such as collections, exception handling, and investor reporting, the advantage shifts to whoever owns the data and workflows rather than whoever performs the work manually.
As a result, many firms are reducing costs while slowly losing their competitive edge. In lending, that shift is already visible. Companies like Upstart built their underwriting on tens of millions of repayment records and now automate most loan decisions for banking partners, giving an advantage to whoever owns the model and the data. A challenger can build a credible operation in eighteen months, while an incumbent may take three to five years to reinvent itself. By the time incumbents recognize the shift, competitors may already have a significant head start.
A better way to start
The answer is not more pilots or a bigger AI budget. It starts with a simple question: will AI change how this business makes money, and who captures the value?
We call this the Value-Locus Test. It helps leaders distinguish between using AI to improve the current business and using AI to reinvent it. A retail mortgage servicer may find that AI changes the economics of its business and requires a different operating model. A relationship-driven private bank may find that AI improves efficiency but does not fundamentally change how it competes. Knowing the difference is one of the most important decisions a leadership team can make.
Once that decision is clear, the focus shifts to two areas: how work is done across people and AI, and the quality of the data that powers those systems.
In regulated industries, trusted data, strong controls, and clear audit trail matters as much as the model themselves. These are the foundations that competitors cannot simply buy.
Why Parkar
This is the work Parkar has been doing in regulated, information-heavy environments for years. We build the data foundations, agentic workflows, and control and audit systems that production AI requires in financial institutions, where every decision must stand up to regulatory scrutiny.
We work closely with servicing and underwriting teams to identify where value is shifting and redesign workflows instead of simply adding AI to existing processes.
The advantage in this wave belongs to incumbents because they already possess what generic models lack: proprietary data and decades of customer trust. Our role is to help turn those assets into a lasting competitive advantage before challengers build their own around them.
Start with an AIONIQ discovery
If your organization is investing in AI today but is not sure whether it is creating new business value or simply making existing processes more efficient, that is worth understanding now rather than later.
The AIONIQ Discovery is a five-day diagnostic that shows where AI can create real business impact, where it cannot, and what it takes to turn AI investments into a lasting competitive advantage.