2026 CMA investigation: What it means for the cloud industry

The UK’s Competition and Markets Authority (CMA) has now set out its latest actions under the Digital Markets Competition Regime (DMCR), following its multi-year Cloud Services Market Investigation. While the regulator has now expanded its focus into business software ecosystems, we must not lose sight of the core issue: the entrenched dominance within the UK's cloud infrastructure.

We have been following this process closely for several years because the implications extend far beyond cloud pricing or procurement practices. What is ultimately being shaped here is the structure of the UK’s digital economy, its dependence on global infrastructure providers, and the degree to which meaningful technological sovereignty can exist in practice.

The CMA has correctly diagnosed many of the structural issues in the market. However, its proposed response raises a more difficult question: whether incremental regulatory engagement is sufficient in a market defined by deep concentration, high switching costs, and vertically integrated control over compute, data, and software ecosystems.

What the CMA investigation has found

The CMA confirms that Amazon Web Services (AWS) and Microsoft Azure together account for approximately 70–80% of the UK cloud infrastructure market, with each holding roughly a 30–40% share individually. Google remains a distant third with a significantly smaller footprint.

This is not a competitive equilibrium in any meaningful sense. It is a duopoly structure reinforced by ecosystem design, pricing models, and technical dependencies that make meaningful switching rare.

The CMA itself estimates that only around 1% of customers switch providers in a given year. This is not a reflection of customer satisfaction, but of structural friction: once workloads are deployed into hyperscaler environments, exit becomes technically complex, financially expensive, and operationally risky.

Key findings

Taken together, these are not isolated inefficiencies. They form a reinforcing system in which technical architecture, pricing models, and software licensing all contribute to the same outcome: reduced ability to switch providers meaningfully, and reflect economic incentives that are unlikely to shift through voluntary commitments alone.

Regulatory progress vs. structural problems 

The CMA has taken a significant step by deciding to launch a Strategic Market Status (SMS) investigation into Microsoft’s business software ecosystem, commencing in May 2026. While this investigation could take up to nine months to conclude, it represents the first time the regulator is using its new Digital Markets powers to directly address the software to cloud pipeline, a structural barrier created by restrictive licensing that Civo has long highlighted as a fundamental threat to fair competition. 

This is an important step forward. However, the regulatory response still leans heavily on phased engagement and voluntary commitments, with an expectation that behavioural change from hyperscalers will gradually address structural issues. That assumption is increasingly difficult to sustain.

The core issue is that many of the proposed remedies rely on cooperation from the same firms whose advantage is built on the constraints being examined. Egress fees, proprietary ecosystems, and licensing advantages are not incidental frictions; they are structural features of the market. Without the enforceable intervention that an SMS designation brings to the entire cloud market, incremental change risks leaving underlying power dynamics intact.

Cloud, AI, and the reality of dependency

Cloud infrastructure is now the foundation layer of AI. Control over compute shapes how artificial intelligence is built and deployed. As a result, concentration in cloud services translates directly into concentration of AI capability influence.

This makes cloud competition a strategic issue, not just a commercial one. While the UK has set out ambitions to lead in AI, supported by initiatives such as the £500m Sovereign AI Fund, the underlying infrastructure remains heavily dependent on a small number of hyperscalers. Public sector procurement decisions further reinforce this dependency, embedding it into critical national systems.

At the same time, organizations are beginning to respond to these risks, increasing diversification and exploring sovereignty-driven strategies. However, rising costs and structural lock-in continue to limit meaningful movement at scale.

What the CMA must do next 

If the UK is serious about addressing structural imbalance in cloud infrastructure, the next phase must move beyond recognition and into timely, enforceable intervention. That means accelerating SMS designations where appropriate, introducing binding requirements on interoperability, and directly addressing the most distortive mechanisms in the market, including egress fees, cloud credits, and licensing asymmetries.

Most importantly, interoperability must become a baseline requirement rather than an optional improvement. Without genuine portability between providers, competition will remain constrained regardless of regulatory intent.

The core issue is not whether the CMA has identified the problem; it has. The issue is whether the response matches the scale of the market failure. For the UK to remain a leader in the age of AI, we cannot afford to be mere tenants in someone else’s digital empire.