For nearly a decade, a cloud-first approach was the default position for most technology leaders in the UK. Boards would green-light public cloud migrations with the promise of agility, rapid deployment and avoiding large capital outlay. Technology teams would deliver, and the business would accelerate.
Yet a quieter, more thoughtful reset is now under way. Across public and private sectors, senior leaders are reassessing what cloud-first truly delivers, and what it costs in practice. This is not a retreat from cloud computing, more a mature examination of where, when and how cloud adds real strategic value. The shift is widespread but subtle, driven by cost pressures, tighter regulation and the complexity of running critical systems at scale.
Cloud-first is colliding with reality
The original pitch for cloud adoption was compelling: defer capital costs, pay for what you use, and scale service delivery as demand requires. But as cloud environments have matured, so too has experience and scepticism.
Finance directors now scrutinise cloud expenditure with the same intensity they apply to energy and property costs, pushing back on unpredictable billing and overruns. It’s not uncommon to hear that cloud environments built quickly have become expensive to run long term; that data transfer fees, duplicated test environments and under-governed usage inflate bills far beyond initial estimates.
A 2024 whitepaper published by Asanti, a cloud service provider, found that the majority (67%) of IT decision-makers wish they had started with a hybrid approach rather than going all-in on public cloud, with many reporting higher-than-expected operating costs. More than half (52%) said they plan to reduce reliance on US public cloud providers, reflecting strategic shifts because of costs and data concerns, while recurring cloud costs were cited by 41% of respondents as a driver for reconsideration of cloud strategy. Other reported drivers included limited control and customisation (39%) and data transfer and latency issues (36%) all factors linked with reassessing public-cloud-only strategies.
These real-world experiences have reframed cloud conversations, too. Where the narrative once focused on how fast workloads could land in the cloud, it now asks where they should reside and why.
Regulation sharpens the focus
In the UK, regulatory requirements add another layer of complexity to cloud strategy. Government and public sector organisations operate under a Cloud First policy introduced in 2013, which directs departments to consider public cloud solutions first when procuring IT services. In practice, that mandate comes with caveats that cloud adoption must be justified with evidence of value for money and suitability. 
For highly regulated sectors such as financial services, health and critical national infrastructure, proving resilience, data sovereignty and uptime isn’t optional. Data residency and compliance requirements mean some systems simply can’t – or shouldn’t – live in shared public environments. The result is a more nuanced conversation on regulatory demands and risk tolerance.
Hybrid is no longer a compromise
‘Hybrid cloud’ used to be shorthand for indecision – a halfway house between legacy infrastructure and public cloud services. Today it is a deliberate architectural choice.
Across UK enterprises, hybrid architectures allow customer-facing or analytics-intensive workloads to benefit from public cloud scale, while stable, predictable core systems remain on private infrastructure where costs are more controlled. That choice isn’t driven by nostalgia for old technology; it’ s driven by economics and governance. Treating infrastructure decisions like portfolio management – balancing speed, cost, control and resilience – has become the new norm. Industry analysts describe this transition as a move from cloud-first to cloud-smart strategy, selecting the best environment for each workload rather than assuming cloud by default.
AI is changing the economics again
Artificial intelligence is forcing another rethink. Training and operating large AI models demands vast compute, high-throughput storage and significant data movement – resources that can push cloud consumption and costs far beyond the lightweight delivery early cloud economics were based on. As a result, finance executives are questioning whether every AI workload truly needs cloud infrastructure, or whether some can be run closer to the enterprise where pricing is more predictable and governance tighter.
This convergence of AI strategy and cloud strategy now regularly features on board agendas, a shift from earlier years when AI was often seen as an application layer rather than an infrastructure cost driver.
Risk and resilience matter more than speed
Speed will always be a competitive advantage – but it no longer dominates the strategic equation. High-profile outages and shared infrastructure failures have reminded leaders that resilience and control matter, particularly where downtime has visible, immediate business impact.
Some organisations are responding by spreading workloads across multiple cloud providers to mitigate shared risks. Others choose to retain mission-critical systems under tighter internal control. Either way, the emphasis has visibly shifted toward assurance, governance and continuity.
This mirrors editorial trends as well: coverage has moved from evangelising early innovation to critically examining how widely accepted strategies perform under real-world constraints.
A quieter and mature phase
Few UK leaders now talk about ‘cloud-first’ and instead terms like ‘cloud-appropriate’ or ‘cloud-smart’ are popping up within LinkedIn thought leadership conversations – language that signals a more thoughtful, workload-centric approach to digital infrastructure.
Cloud computing in the UK isn’t unwinding or being disrupted it’s maturing. The initial excitement has given way to experienced practice, and leaders still see value in cloud services. But they are treating cloud spending, governance and resilience with the same strategic rigor applied to other business assets.
These real-world experiences have reframed cloud conversations, too. Where the narrative once focused on how fast workloads could land in the cloud, it now asks where they should reside and why.
For nearly a decade, a cloud-first approach was the default position for most technology leaders in the UK. Boards would green-light public cloud migrations with the promise of agility, rapid deployment and avoiding large capital outlay. Technology teams would deliver, and the business would accelerate.
Yet a quieter, more thoughtful reset is now under way. Across public and private sectors, senior leaders are reassessing what cloud-first truly delivers, and what it costs in practice. This is not a retreat from cloud computing, more a mature examination of where, when and how cloud adds real strategic value. The shift is widespread but subtle, driven by cost pressures, tighter regulation and the complexity of running critical systems at scale.
