Balancing cloud agility and on-premises control for IT resilience

For almost 20 years, cloud computing has been the foundational engine of modern business transformation, delivering the essential scalability and agility that fuels business growth. Today, a cloud-first approach is the standard for deploying applications and managing data, contributing to the $723bn+ end-user spending as of 2025, according to Gartner.

However, as AI adoption increases, the demands on our digital infrastructure are changing once again. AI is rapidly reshaping cloud architecture and economics, driving infrastructure demand that requires a more sophisticated approach to workload placement. This isn’t a move away from the cloud, but a move towards unlocking its full potential by integrating it into a more holistic, balanced hybrid strategy.

By blending the unparalleled agility of public cloud with the dedicated performance and control of on-premises infrastructure, organisations can build a truly optimised and resilient IT estate ready for the demands of AI.

A strategic evolution in data placement

Understanding this strategic evolution requires looking back at how we got here. The initial migration to the cloud was driven by clear and compelling business advantages. Organisations across every industry embraced the promise of enhanced scalability, cost-efficiency, and accessibility. This shift was instrumental in supporting a global, distributed workforce, perhaps most notably during the pandemic, when healthcare organisations relied on cloud platforms to share critical patient information across dispersed facilities.

Today, after that initial ‘all-in’ approach, we are seeing a shift towards hybrid models that blend cloud and on-premises solutions. This evolution is being driven by the demands of applications like AI, a trend confirmed by recent GTT research showing that more than half (51%) of all AI workloads are now running in private cloud or on-premise environments. This move requires greater control over sensitive data and tighter integration with core business processes. And as the stakes get higher, the conversation around data sovereignty is intensifying.

The goal is no longer simply to be ‘in the cloud’, but to have a clear, unified strategy that governs where data is stored and who has access to it, allowing organisations to harness the agility of the cloud while maintaining the robust performance and governance of on-premises infrastructure.

From data governance to customer trust

This need for robust performance and governance is critical, especially as credential abuse continues to be a leading initial attack vector, as highlighted in a 2025 Verizon Business report. The

consequences of exposure can be significant, ranging from disruptive ransomware attacks to regulatory fines and the erosion of customer trust.

This is precisely why we are seeing organisations in data-sensitive sectors, such as financial services, strategically investing in their on-premises infrastructure. Barclays, for example, has migrated more than 50,000 workloads to a private cloud infrastructure. This move enables them to support demanding AI initiatives with high-density compute solutions, while maintaining stricter control over their data to meet security and compliance obligations. It’s a perfect example of using the right tool for the right job – marrying the security of on-premises infrastructure with the agility of the cloud across the wider business.

Building resilience beyond the digital realm

A truly comprehensive IT strategy must extend beyond digital threats and consider the physical infrastructure that powers our services. Cloud data centres are marvels of engineering, but their operations are dependent on vast amounts of energy and water. This reliance on physical resources introduces a new set of strategic considerations, as natural resource volatility can create unexpected points of failure that impact service availability.

This physical infrastructure dependency creates a business continuity challenge that many organisations overlook – one that can be as damaging as a cyber breach. As resource challenges intensify, organisations that rely exclusively on a single cloud model may face unforeseen availability and cost issues. This is why many industry leaders are actively seeking alternative power sources for their data centres to ensure reliable and sustainable operations. Google’s deal to use small nuclear reactors to power its AI data centres is a prime example of this forward-thinking approach.

By considering the physical dependencies of their digital services, businesses can better maintain operational continuity and cost predictability for key workloads, strengthening their entire IT estate against future shocks.

A unified path forward

How can an organisation strengthen its estate against both digital and physical shocks? The path forward lies in a unified IT infrastructure that harmonises the benefits of cloud and on-premises solutions. This isn’t about acquiring more tools, but adopting a new mindset that provides a single, holistic view of the entire data estate. An effective strategy must be forward-looking, implementing a consistent governance framework that dictates how information is managed, whether it’s in a public cloud or a private data centre.

Crucially, this must also account for long-term resilience by factoring in physical risks like resource volatility alongside digital threats. The key is not choosing between cloud or on-premises, but

creating an integrated strategy where both environments work together seamlessly. By doing so, organisations can achieve optimal security and resilience against both cyber and physical challenges, ensuring they are perfectly positioned to meet evolving customer needs and capture the full value of the AI era.