In chaos theory, a butterfly flapping its wings can trigger a hurricane on the other side of the world. While hyperbolic, the past few weeks exemplified how a small change can create massive global waves.
Anyone who watched the markets in early 2025 when DeepSeek—a relatively unknown player in the AI space—announced they had built something remarkable. They’ve built an AI model that could match the capabilities of tech giants who had spent billions—but at a fraction of the cost. The announcement wiped 3% off the Nasdaq in a single day, and removed $600m in value from Nvidia alone. Before the ripples could settle, Alibaba claimed their models could outperform DeepSeek’s, sending fresh waves through the market.
The LLM shakeup, coupled with the recent announcement of 25% tariffs on advanced chips, exemplifies the challenges banks face in 2025. How do you maintain resilience and customer focus when technology and market dynamics shift dramatically overnight?
Three pressures reshaping banking priorities
Banks are grappling with three interconnected pressure points reshaping their operational landscape.
First, customer expectations continue to evolve beyond traditional personalization. Today’s clients demand products that align not just with their life stage but with their precise moment in life—all while banks face mounting pressure to optimize efficiency and manage costs as rate cuts look slow or entirely unlikely. With the headwinds of slow rate cuts in the US and continued inflation, balancing these two priorities will be a challenge that requires new approaches.
Second, we’re witnessing a significant divergence in regulatory approaches to operational resilience. While Europe implements increasingly structured oversight through initiatives like the EU’s AI Act, the United States’ new administration is moving toward a model of self-governance. This regulatory bifurcation creates unique challenges for global institutions that must maintain operational consistency across different jurisdictions.
Third, the technology infrastructure landscape is becoming more complex and costly. The introduction of 25% tariffs on crucial components like chips affects everything from AI deployment costs to Cloud infrastructure expenses, eroding efficiency gains that banks were counting on from their AI investments.
Building a resilient and adaptable foundation
The path forward requires a balanced approach to operational transformation. The answer lies in developing what might be called a “customer-cost-centric transformation” approach. This isn’t just about balancing customer experience with efficiency—it’s about fundamentally reimagining how banks deliver value in an AI-driven world.
A recent incident at a major bank illustrates the challenges. An executive assistant used an internal AI copilot to summarize confidential board discussions. Yet they only realized later that the summary would be accessible to all employees—including interns.
This near-miss with potential insider trading exposure underscores a crucial point with the rapid pace of AI deployments. Responsible AI implementation requires sophisticated governance frameworks that balance capability with control. Focusing too much on either side will ultimately lead to an equal or greater risk to organizations.
The path forward: Three strategic imperatives
Create a cohesive, open ecosystem
The evolution from open banking to open finance, and now to a future of an “open economy,” requires a more sophisticated approach than basic connectivity. Banks must have an ecosystem mindset and create a platform approach to handle the complexity and increasing scale. They also need to be flexible and composable enough to respect increasingly distinct regional boundaries while enabling collaboration and iterative innovation.
This means building a platform that enables the reusability of infrastructure and tech investments while operating differently in various jurisdictions. For example, supporting specific API and data sharing policies in regions like Europe while enabling more controlled data exchange in others.
The goal isn’t universal openness but a strategic platform approach that helps banks grow and delight customers while improving costs to serve them.
Blend deterministic and non-deterministic processes
Many firms are stuck “planting 1000 flowers” of AI experiments trying to find what brings the most value. Banks must clearly differentiate between operations that require consistent, deterministic outcomes and those where AI can provide valuable supplementary support. This distinction helps optimize investments while maintaining appropriate control over AI deployments and how they operate.
The flexibility to blend agentic AI with deterministic flows enables firms to increase their overall automation rate and free capacity so employees can work on higher-value—and ultimately—more human-centric activities. Simply put, you can model the portion of your process that requires predictability and control. Then let AI agents handle work that is more dynamic in nature.
Build enterprise-scale deployment models
As AI becomes more deeply embedded in banking operations, governance frameworks must evolve beyond basic risk management to comprehensive operational controls. This includes everything from data access protocols to model deployment strategies and ethical considerations.
Additionally, they need the flexibility to swap out models with ones that offer new value, whether that’s lower cost or improvements in accuracy or speed. Balancing these opposing forces will be a continuous challenge, reinforcing the need for adaptability and composability.
Looking ahead
The successful banks of tomorrow will be those that can build technology platforms capable of absorbing market shocks while keeping the business moving forward. While technology is an important ingredient, it demands a fundamental rethinking of operational excellence and business resilience
As we progress through 2025, banks need to focus on developing flexible, agile process platforms—systems and processes that can evolve without compromising stability or compliance. This means investing in infrastructure that supports both current needs and future possibilities while maintaining the right balance of governance and freedom to innovate.
The challenge isn’t predicting every potential market disruption. That’s nearly impossible (even if we were three for three in this episode!). It’s building a technology foundation that can adapt to unexpected changes while delivering value today and sustainably into the future.
For banks that get this right, the reward will be the ability to turn future tsunamis into surfable waves, maintaining operational excellence even as the storms rage on.
Want to see how these trends are playing out against the original predictions? Read Financial Services Industry Outlook & Guidance for 2025 for a deeper look at Sathya’s predictions for what will shape the industry in the coming year.
Explore Camunda for Banking and Financial Services to learn more about how leading banks are building resilient platforms to operationalize AI and iterative innovation.
Start the discussion at forum.camunda.io