CMA Targets Microsoft’s Business Software Ecosystem in UK Cloud Antitrust Probe

Microsoft’s latest UK antitrust problem is not that regulators are widening their cloud review, but that they are zeroing in on the business software layer that feeds the cloud stack beneath it. The Competition and Markets Authority has now moved toward a strategic market status investigation into Microsoft’s business software ecosystem, a shift that appears designed to tackle cloud licensing and bundling issues through a narrower but more actionable lens. That change matters because the CMA has already spent more than two years studying the cloud market and concluded that competition is not working well enough; now it is choosing the path most likely to produce remedies rather than another broad diagnostic report.

Background​

The UK cloud probe did not begin with Microsoft alone, but Microsoft quickly became one of its central targets. The CMA’s cloud-services market investigation grew out of earlier Ofcom work and focused on the structure of the public-cloud market, where Microsoft Azure and Amazon Web Services dominate spending and where switching costs, egress fees and software licensing practices have long troubled customers and rivals. By July 2025, the CMA’s final decision had landed on a familiar but still forceful conclusion: competition in UK cloud services was not functioning as well as it should.
That final decision was notable not because it produced immediate remedies, but because it pushed the CMA toward a more modern set of powers under the Digital Markets, Competition and Consumers Act 2024. Rather than trying to fix cloud competition with a single market-investigation remedy package, the regulator recommended strategic market status investigations into the largest providers. The logic was simple: if the market is structurally tilted, the CMA’s digital-markets toolkit may be more effective than a conventional competition review.
Microsoft’s role has always been broader than a single cloud product. The company’s software stack spans productivity software, identity services, device management, collaboration tools, developer tooling and cloud infrastructure. That gives it unusual leverage because a business can buy software from Microsoft, authenticate users through Microsoft, manage devices through Microsoft and then deploy workloads on Azure. In practice, the cloud market is not just about infrastructure capacity; it is also about how software licensing and ecosystem design shape where customers feel they are allowed to run their workloads.
The new investigation track reflects a regulatory realization that cloud scrutiny alone may be too blunt. If the real leverage point is business software licensing, then the cloud inquiry needs to flow through that gateway. That is why the CMA’s reported decision to narrow its focus looks less like a retreat and more like a tactical recalibration. It is a shift from surveying the whole battlefield to targeting the terrain where Microsoft’s power is most likely to distort downstream competition.

Why the timeline matters​

The timing is especially important because the CMA’s new regime only came into force in 2025, and the regulator has been building a portfolio of designated digital giants. Google has already been confirmed with strategic market status in search services, and Apple and Google were both designated in mobile platforms after separate SMS investigations. That gives the CMA a practical blueprint for how a Microsoft probe could unfold: designate first, then consult on conduct requirements and pro-competition interventions later.

Why Microsoft’s business software matters more than the cloud label​

The phrase “business software ecosystem” sounds broader, and that is the point. Microsoft is not merely being examined as a cloud vendor; it is being examined as a platform owner whose software choices can tilt customer behavior toward or away from Azure and away from competing clouds. That ecosystem includes licensing terms, interoperability choices and product packaging decisions that may make rival clouds more expensive or less attractive to use.
This matters because cloud competition rarely fails in just one place. Sometimes the headline price of compute looks competitive, but the surrounding ecosystem makes migration costly. Customers can face higher charges, more restrictions or more operational friction when they try to run Microsoft software on a non-Microsoft cloud. The CMA’s past work suggested that those frictions were not incidental; they were part of the market structure it was trying to address.

Licensing as leverage​

Microsoft’s licensing practices have been the central recurring complaint. Rival providers and customers have argued that Microsoft makes it less attractive to use Microsoft software on competing infrastructure than on Azure, creating an artificial advantage for its own cloud. That is a classic competition-law concern: not just that a firm is big, but that its commercial terms shape the terms of rivalry itself.
The practical significance is easy to miss unless you run enterprise IT at scale. A software license that seems minor in isolation can become decisive when multiplied across thousands of users, workloads and compliance obligations. In that setting, small pricing asymmetries can alter architecture decisions, cloud-provider selection and long-term migration plans. The CMA’s focus suggests it sees exactly that kind of cumulative effect.
Key implications include:

• Licensing terms can function like a hidden switching cost.
• Bundled software ecosystems can make a rival cloud less usable.
• Enterprise procurement teams may face fewer genuine choices than price sheets suggest.
• Public-sector buyers may be especially exposed because of legacy Microsoft dependence.
• Regulatory scrutiny can now target conduct, not just market concentration.

The CMA’s narrower lens is also a stronger enforcement lens​

The move from a market investigation to a strategic-market-status probe is not just procedural. It changes the regulator’s posture from diagnosing structural issues to potentially imposing targeted obligations on a designated firm. That is why the new cloud scrutiny looks narrower even as it becomes more intense. The CMA is effectively saying it already knows where the problem lives; now it wants the legal tools to operate on that exact point.
This is also more politically durable. A broad cloud-market remedy package could be contested as overreach or as too difficult to administer across many providers. A targeted Microsoft SMS investigation, by contrast, can be framed as a focused response to identifiable conduct with measurable effects. That makes the process more likely to survive both legal challenge and industry pushback.

From market study to conduct regulation​

The CMA’s own material shows a transition in policy architecture. Its cloud investigation concluded with a recommendation to use digital-markets powers, and the regulator said in June 2025 that it expected the board to consider further designation options in early 2026. That is exactly the kind of sequencing that would lead to a business-software probe as a precursor to conduct rules on licensing or interoperability.
In regulatory terms, that means the most important outcome may not be a formal finding of wrongdoing. It may be the opening of a legal pathway that lets the CMA impose guardrails on how Microsoft packages, prices and interoperates across its ecosystem. For a company with such a sprawling enterprise footprint, that is potentially more consequential than a one-off fine.

AI is now part of the cloud competition story​

It would be a mistake to read this as an old-fashioned licensing case with some cloud jargon attached. AI has become a central reason regulators are paying closer attention to the cloud stack, because modern AI services sit on top of cloud infrastructure, enterprise software, identity systems and developer tools all at once. If Microsoft’s ecosystem controls access to that stack, then AI competition can be affected indirectly even before a model is trained or a chatbot is launched.
The MLex description is especially revealing in suggesting that regulators expect “noise” over AI’s role in the cloud ecosystem and the UK economy. That is not surprising. AI workloads intensify demand for compute, storage, data transfer and enterprise integration, which makes the cloud provider question more strategic than ever. In that context, a business-software probe is not just about licensing; it is about who gets to shape the next generation of digital infrastructure.

Why AI amplifies cloud power​

AI increases the value of platform integration. A vendor that already controls productivity tools, identity, hosting and developer tooling can more easily package AI features into daily workflows, and customers may find it difficult to disentangle one layer from another. That creates a networked advantage that is harder to unwind than a simple market-share problem.
It also makes regulator behavior more urgent. If AI-related enterprise adoption becomes tightly bound to one ecosystem, the downside is not just higher costs. It can mean less portability, less bargaining power and fewer chances for smaller providers to compete on merit. The CMA appears to be testing whether Microsoft’s ecosystem is becoming one of those bottlenecks.

Enterprise customers are the real audience of this probe​

For consumers, this is an abstract competition case about technology giants and market structure. For enterprises, it is much more concrete. UK businesses often build their IT estates around Microsoft identities, Microsoft licensing, Microsoft endpoint management and Microsoft collaboration tools; if those decisions lock them into a preferred cloud path, then procurement flexibility is already compromised before cloud negotiations begin.
That is why the CMA repeatedly links this work to productivity and value for money across the private and public sector. The regulator’s concern is not simply that Microsoft is dominant. It is that dominance in software can spill over into cloud, and that cloud can in turn shape overall enterprise costs, resilience and innovation capacity. Those effects accumulate quietly over years, which is why competition agencies are often slower than technology markets themselves.

Enterprise pain points to watch​

The most likely flashpoints are not flashy consumer issues but dull, expensive ones. Licensing discounts, migration penalties, identity dependencies and cross-platform support terms are the kind of issues that can trap an IT department even when no one individual contract looks abusive. The CMA’s narrowed scrutiny suggests that these small frictions are exactly what it wants to test.
At a practical level, enterprise buyers should expect more attention to:

• Hybrid-cloud portability
• Licensing parity across clouds
• Identity and management interoperability
• Discount structures tied to bundled commitments
• Migration costs and exit friction
• Procurement transparency in public and private sectors

The competitive consequences for AWS and Google are not symmetrical​

A Microsoft-focused probe does not automatically mean AWS escapes scrutiny in a broader sense, but it does suggest the CMA believes Microsoft’s software ecosystem creates a distinct kind of competitive issue. AWS has been part of the cloud investigation and remains relevant to the market’s concentration, yet Microsoft’s leverage comes from a different source: software ownership that touches the operating environment itself.
That distinction matters because it may influence future remedies. AWS’s issues are likely to revolve more around cloud market concentration, while Microsoft’s may revolve around whether enterprise software terms improperly steer demand toward Azure or disadvantage rival clouds. If so, the regulator may end up writing two different competition narratives for two cloud giants operating in the same market.

A more precise remedy map​

For rivals, the upside is obvious: clearer rules could reduce the burden of competing against a vertically integrated incumbent. If Microsoft cannot use software licensing to tilt cloud demand, the market could become more contestable and pricing pressure could improve. That would be especially helpful for smaller infrastructure providers and managed-service vendors trying to win enterprise workloads.
But the regulatory map is also more complicated than simply “help the underdogs.” If the CMA intervenes too aggressively, it could create uncertainty about legitimate bundling, security integration and product support. That is why the regulator’s targeted approach matters: it suggests an attempt to separate harmful lock-in from ordinary product integration.

The political economy of the probe is just as important as the law​

The UK government has made growth, investment and productivity central to its industrial agenda, and the CMA has been operating under a stronger political expectation that competition enforcement should support economic dynamism. That does not mean regulators will go easy on Microsoft; it means they will likely try to frame interventions as pro-growth rather than anti-business.
That framing is important because cloud infrastructure is now seen as national economic plumbing. If the software ecosystem controlling cloud usage is distorted, the effects are not confined to a single corporate balance sheet. They spread through startups, government IT contracts, public procurement, AI adoption and the broader cost structure of digital transformation. That is why this investigation has the potential to become one of the CMA’s most consequential digital-markets tests.

A regulator learning to use its new tools​

The CMA’s recent actions show a pattern. It has designated Google with SMS in search, designated Apple and Google in mobile platforms, and used consultation-heavy roadmaps to explain how interventions might follow. Microsoft now appears to be next in line, but in a different domain where the conduct concerns are less about apps and more about enterprise software economics.
That sequence is not accidental. It reflects a regulator trying to build precedent, credibility and procedural discipline around a brand-new regime. The Microsoft probe will therefore be watched not just for what it says about cloud licensing, but for how the CMA operationalizes its broader digital-markets powers.

Strengths and Opportunities​

The strongest argument for the CMA’s narrower Microsoft scrutiny is that it aligns regulatory effort with the market’s actual bottleneck. If software licensing is the lever that distorts cloud competition, then targeting that lever is more efficient than pursuing a generic cloud fix. That is why the investigation could produce remedies with more real-world impact than a broader but fuzzier inquiry. The chance to improve portability, lower switching costs and strengthen buyer leverage is significant.

• More precise enforcement than a broad market study
• Potential relief for enterprise customers facing hidden lock-in
• Better cloud portability if licensing parity improves
• Stronger competition for rival infrastructure providers
• Clearer rules for software-cloud bundling
• A useful precedent for the CMA’s digital-markets regime
• Possible benefits for AI competition through less ecosystem friction

Risks and Concerns​

The biggest risk is that the probe overshoots and treats integration itself as anti-competitive. Microsoft’s ecosystem is powerful partly because it is deeply connected, and some of that integration delivers real security, manageability and user-value benefits. If the CMA draws the line too broadly, it could chill legitimate product design or create compliance complexity that hurts customers as much as incumbents. The challenge will be distinguishing harmful lock-in from ordinary platform coherence.

• Overbroad remedies could disrupt legitimate bundling
• Legal challenge may slow or narrow any interventions
• Uncertainty for enterprise procurement during the probe
• Compliance burden for Microsoft and possibly partners
• Risk of unintended effects on security and interoperability
• Potential duplication with EU or other international actions
• Regulatory fatigue if multiple overlapping probes emerge

Looking Ahead​

The next phase will likely determine whether this becomes a landmark conduct case or just another expensive investigation. Much will depend on whether the CMA can translate its cloud findings into concrete questions about Microsoft’s business software terms, rather than drifting into a generalized critique of size and success. If it succeeds, the case could reshape how licensing, interoperability and cloud access are governed in the UK for years to come.
Expect the debate to intensify around AI, too. The UK’s regulatory community is increasingly aware that control over enterprise software and cloud infrastructure may become a proxy battle for influence over AI distribution and commercial adoption. If Microsoft’s ecosystem is found to create friction for rival cloud use, the implications will extend well beyond cloud procurement and into the architecture of the AI economy itself.

• The SMS investigation launch date and scope definition
• Whether licensing is formally singled out as a competition concern
• Any consultation on conduct requirements or pro-competition interventions
• Responses from enterprise customers and cloud rivals
• Signs of coordination or contrast with EU and US antitrust thinking

The broader lesson is that antitrust scrutiny of Big Tech is becoming more modular and more technically specific. Microsoft is no longer being viewed only as a cloud provider, or only as a software vendor, or only as an AI enabler. It is being examined as an ecosystem owner whose leverage spans all three, and that is precisely why the UK regulator seems determined to narrow in rather than scatter its fire.

---

Source: MLex Microsoft software probe sees UK competition regulator narrow its cloud scrutiny | MLex | Specialist news and analysis on legal risk and regulation

 

[ChatGPT]

Microsoft’s next regulatory headache in the UK is shaping up to be less about raw cloud infrastructure and more about the glue that holds enterprise software stacks together. The Competition and Markets Authority is reportedly preparing a Strategic Market Status probe that could put Windows, Microsoft 365, and Copilot under the microscope, with regulators increasingly focused on whether Microsoft’s licensing and product bundling give it an unfair edge across cloud and productivity markets. That matters because the UK already concluded that cloud competition is not functioning as well as it should, and it is now moving from diagnosis to remedies. In plain English: this is no longer just about where companies host data; it is about how Microsoft can influence where they buy software, how they move workloads, and how easily they can switch suppliers.

Overview​

The CMA’s interest in Microsoft did not appear from nowhere. It follows a long-running cloud services market investigation that began after Ofcom referred the sector, and that inquiry has already provisionally found that competition is not working as well as it should in the UK’s roughly £9 billion cloud services market. The regulator’s concern has been consistent: high switching costs, technical barriers, and software licensing practices may be making it harder for businesses to move between providers or build genuinely multi-cloud environments.
That broader cloud probe is important because it provides the policy runway for the next phase. The CMA has explicitly said it may use its new digital markets powers to consider whether the biggest providers, including Microsoft, should receive SMS designation in cloud services. It also said potential interventions could include technical standardisation, lower data transfer charges, and fairer software licensing. The new reported software-focused investigation looks like an extension of that logic rather than a separate regulatory universe.
The timing is also telling. The CMA has been building a broader toolkit under the Digital Markets, Competition and Consumers Act 2024, and it has already applied strategic market status to mobile platforms in another sector. That signals a regulator that is increasingly comfortable using ex ante powers instead of relying only on traditional competition cases after harm has already calcified. Microsoft is therefore facing a regulator that is not just asking whether it has market power, but whether its business architecture itself distorts choice.
Copilot is the obvious modern twist. Earlier cloud debates centered on server licensing, egress fees, and workload portability. Now the issue is whether a dominant productivity and AI suite can make it harder for rivals to compete on the desktop, in collaboration software, and in emerging AI assistants. That is a more complicated problem, because it combines classic antitrust themes with a much newer question: who gets to control the front door to work itself?

---

Why the CMA Is Looking Beyond Cloud Infrastructure​

The immediate catalyst is not just infrastructure pricing. The CMA has already spent months examining Microsoft’s software licensing practices in the cloud context, and those papers and submissions repeatedly return to the same theme: Microsoft can make it harder for customers to run its software on non-Microsoft clouds, and that can shape purchasing behavior upstream. The regulator’s concern is therefore not only about Azure versus AWS or Google Cloud, but about the way server software and productivity tools can influence cloud competition.
That matters because enterprise customers do not buy cloud services in isolation. They buy Windows Server, SQL Server, Microsoft 365, identity tools, endpoint management, and now Copilot as part of a larger stack. When one vendor controls both the workload layer and the productivity layer, switching becomes a systems problem rather than a procurement decision. The CMA appears to be asking whether Microsoft can use that layered position to steer customers into its own ecosystem.

The licensing question is the real lever​

The strongest regulatory pressure point may be licensing rather than cloud capacity. The cloud investigation’s working materials and third-party responses repeatedly describe Microsoft’s licensing as a structural barrier, especially where customers want to use Microsoft software on rival clouds. In a market where price transparency is already poor, licensing complexity can work like a hidden toll booth.
That is why fair licensing appears so often in the CMA’s language. The regulator has already said interventions might ensure fair licensing of software, and it is hard to imagine a future SMS regime that ignores the way software terms can reinforce cloud lock-in. If Microsoft’s licensing rules change the economics of where software runs, then competition in the cloud market becomes partly a licensing problem in disguise.

• The CMA has already tied cloud competition concerns to software licensing.
• Licensing terms can influence where workloads are deployed.
• Software and cloud are increasingly one market from a customer’s perspective.
• Regulators are treating portability as a competition issue, not just an IT feature.

Why this extends beyond Azure​

The strategic importance of the new probe is that it reaches into areas where Microsoft has been especially successful: productivity software and AI integration. If the regulator looks at Windows, Office, and Copilot together, it is effectively examining whether Microsoft can bundle the “work surface” with the cloud back end in a way competitors cannot match. That is a far more ambitious theory of harm than a simple cloud price dispute.
For rivals, that creates an opportunity and a challenge. Cloud providers have long argued that customers are not comparing clouds on a level playing field because Microsoft can offer software economics that others cannot replicate. But a remedy that reaches into office software and AI assistants could also pull the CMA into a much larger debate about how much integration consumers actually want versus how much competition law should permit.

---

Copilot Changes the Antitrust Conversation​

Copilot changes the temperature of this story because it turns a software licensing dispute into a contest over the future interface to work. If Microsoft can embed AI deeply into Office, Windows, and collaboration tools, it may create a powerful default effect that rivals cannot easily overcome. Regulators are likely to ask whether that default is the result of product quality, or whether it is reinforced by commercial design choices that reduce meaningful alternatives.
The CMA’s concern seems to be less about the existence of Copilot itself and more about whether Microsoft can use it to intensify ecosystem lock-in. If a customer already depends on Microsoft 365, Entra-style identity, and Azure-linked management tools, adding AI features inside that stack may raise the cost of moving to another vendor. That is precisely the kind of “bundling plus switching friction” combination that competition authorities now view as especially potent.

The enterprise case​

Enterprises are unlikely to object to features that save time. Indeed, Microsoft has been able to point to strong internal and public-sector adoption cases for Copilot, including productivity claims in UK government and NHS contexts. But regulators are not evaluating only efficiency gains; they are evaluating market structure, and efficiency can coexist with exclusion if customers face fewer real choices over time.
The enterprise risk is that AI becomes another layer of dependency rather than a source of flexibility. If procurement teams cannot easily compare Microsoft’s AI suite with competing tools because switching would ripple through document workflows, security settings, and licensing terms, then “best product wins” becomes a less credible story. That is why Copilot is not a side issue here; it is central to the regulator’s theory of modern lock-in.

• Copilot could increase Microsoft’s default advantage.
• AI features may deepen switching costs inside enterprise workflows.
• Regulators may treat bundled AI as a competition issue, not just a product feature.
• Procurement teams may face more complexity, not less, if AI is tied tightly to platform licensing.

The consumer spillover​

Consumer users are less likely to care about formal licensing structures, but they will still feel the effects if enterprise defaults shape the tools they use at work. Microsoft’s broad suite integration can normalize a single-vendor environment across personal and professional computing, making alternatives feel unfamiliar or risky. Over time, that kind of soft lock-in can matter as much as any contract clause. That is the subtle antitrust problem here.
The competitive question is whether AI will widen the gap between incumbents and challengers. If a company can bundle AI into productivity software that is already embedded in business processes, then new entrants must compete not just on model quality but on workflow displacement, admin overhead, and trust. That is a very high bar.

---

Cloud Switching Costs Are Still the Core Problem​

Even with the focus shifting toward software, the cloud issue remains foundational. The CMA has already spent significant time on egress fees, interoperability, and technical barriers to switching. Those problems remain especially relevant because the cloud market is not sold as a single product; it is sold as an ecosystem of storage, networking, databases, identity, security, and management services that are difficult to unwind.
Microsoft has responded by emphasizing that it has made progress on switching and interoperability, including removing egress fees in some contexts and offering free transfer windows. The company has also said it agrees with the principles of moving workloads freely and avoiding vendor lock-in. That posture may help politically, but it does not end the regulatory question, because the CMA can still ask whether the broader commercial structure remains anti-competitive even if one fee is reduced.

Egress fees are only part of the bill​

Egress fees are visible, which makes them politically useful, but they are not the only cost of moving. Licensing rules, data gravity, engineering rework, staff retraining, and architectural redesign can all make migration expensive long after a transfer fee disappears. That is why the CMA’s analysis has repeatedly linked technical barriers and software rules; the regulator appears to understand that lock-in can be embedded in the architecture itself.
There is also a strategic angle for Microsoft. By promising to improve interoperability while defending the rest of its stack, it can present itself as reform-minded without surrendering the commercial advantages of integrated software. That may be sufficient in some policy circles, but regulators may increasingly prefer hard obligations over voluntary commitments.

• Switching costs include more than data transfer fees.
• Licensing and engineering effort can be more important than headline pricing.
• Interoperability improvements may reduce friction without fully fixing market power.
• Voluntary commitments often look better than they function in practice.

Multi-cloud is the benchmark, not the exception​

The CMA’s focus on multi-cloud reflects a broader market shift. Large organizations increasingly want resilience, bargaining power, and workload flexibility rather than total dependence on one supplier. If Microsoft’s terms make multi-cloud materially harder, that may become a competition issue in its own right because it constrains the procurement models that sophisticated customers are now trying to adopt.
That has knock-on effects for the entire ecosystem. Independent providers such as Civo, OVHcloud, and others have a clear interest in regulatory action that reduces the gravitational pull of the hyperscalers. Their argument is not simply that Microsoft is large; it is that Microsoft’s scale can become self-reinforcing if software, identity, and cloud infrastructure are too tightly interwoven.

---

What Microsoft Is Saying​

Microsoft’s public posture is broadly cooperative. Brad Smith has said the company is committed to working quickly and constructively with the CMA and providing the information needed for its reviews. That is classic Microsoft regulatory language: respectful, future-focused, and designed to signal that the company is part of the solution rather than the problem.
The company also argues that it has already made meaningful changes. Microsoft says it has abolished egress fees globally for switching providers, extended its free switching window to 180 days, and allows free data transfers. Whether those measures satisfy regulators is a different matter, but they do show that Microsoft understands the political importance of portability and the reputational damage of being seen as a lock-in architect.

The defense is now about principles​

Microsoft has framed the issue around principles like portability, interoperability, and customer choice. That is smart because it accepts the language of the debate without conceding the legal conclusion. It also lets the company argue that market outcomes should be judged by customer demand, not by whether a regulator dislikes the degree of integration on offer.
But a principles-based defense has limits when the regulator is investigating market structure. The CMA does not have to prove that Microsoft is doing something obviously malicious; it only has to decide whether market features are preventing, restricting, or distorting competition. In that frame, even a well-intentioned integrated strategy can still be a competition problem.

• Microsoft is leaning on interoperability language.
• It is emphasizing concrete changes like lower switching friction.
• The company is trying to show cooperation rather than confrontation.
• Regulators may still see structural harm even if Microsoft’s intentions are benign.

Why Brad Smith matters here​

Smith’s role is especially significant because Microsoft has long used him as the public face of its regulatory diplomacy. His messaging tends to blend legal caution with a broader narrative that Microsoft supports competition, openness, and customer control. That can be persuasive, but it also means the company is being judged against a long-standing promise of responsible platform stewardship.
This is where history becomes awkward for Microsoft. The company has spent years arguing that it embraces interoperability and standards, yet regulators and rivals still see recurring patterns of dependency, especially when enterprise software and cloud services are combined. The new probe suggests that those old concerns have not gone away; they have simply evolved into a more AI-shaped version of the same debate.

---

The Competitive Implications for Rivals​

For rivals, the upside of a successful CMA probe is obvious: more room to compete on price, service, and migration friendliness. If the regulator forces clearer licensing rules or more neutral software treatment across clouds, then Microsoft’s competitors may find it easier to win enterprise workloads that are currently held in place by contract inertia. That could be especially important for smaller UK and European providers that have struggled to overcome hyperscaler scale.
The downside is more nuanced. A heavy-handed remedy could create compliance complexity that only the largest companies can manage, which would ironically favor the same giants regulators are trying to constrain. If the CMA wants to help smaller providers, it will need remedies that are simple enough to be enforced and robust enough to change behavior. That is a difficult balance.

Google and AWS are not neutral observers​

Google has been particularly vocal in the UK cloud probe, backing the CMA’s concerns about Microsoft’s licensing practices. That does not make Google disinterested; it makes Google a market rival with obvious commercial incentives to support remedies that reduce Microsoft’s advantage. Still, the presence of a serious competitor pressuring for change can reinforce the regulator’s sense that the market is structurally skewed.
AWS, meanwhile, is also in the regulatory frame. The CMA’s cloud work has treated Microsoft and AWS as the two largest providers, and the regulator has indicated that both may merit SMS consideration. So while Microsoft is in the headline, the larger story is a broader reevaluation of how hyperscale cloud power works across the UK economy.

• Smaller providers want a more level playing field.
• Bigger rivals may welcome rules that narrow Microsoft’s bundling advantage.
• The CMA must avoid remedies that only the largest firms can absorb.
• Competition in cloud is increasingly a policy issue, not just a market outcome.

The market is moving faster than the law​

One reason this probe feels urgent is that AI adoption is changing enterprise buying patterns faster than traditional regulation can respond. Microsoft has an opportunity to fold AI into existing enterprise agreements, while smaller vendors must persuade buyers to add yet another tool to an already crowded stack. The longer that imbalance persists, the harder it becomes to unwind.
That is why the CMA’s digital markets powers matter. They are meant to intervene earlier, before competition is fully foreclosed. If the regulator waits until AI productivity software has become as entrenched as email or identity management, the remedies may arrive too late to meaningfully change the market.

---

The Likely Remedy Landscape​

If the CMA proceeds, the remedy menu could be broad. Based on its prior cloud findings, it may consider obligations on interoperability, software licensing transparency, data portability, and rules that prevent favoritism toward Microsoft’s own cloud when its software is deployed elsewhere. The regulator has already hinted that it sees value in measures that help customers switch or use multi-cloud models more freely.
That said, remedies are never as neat in practice as they sound on paper. The real challenge is to write obligations that are specific enough to be enforceable but general enough to survive changes in product architecture. With AI evolving rapidly, the CMA risks creating a rulebook that is obsolete before it is fully implemented.

Conduct remedies versus structural fixes​

The most likely first step is conduct-based intervention rather than structural separation. Regulators usually prefer behavior rules because they are easier to justify and less disruptive than breaking up businesses. But conduct remedies also demand constant supervision, and they can degenerate into a compliance treadmill if the underlying incentives remain unchanged.
Structural remedies would be far more controversial and much less likely at this stage. Still, even the possibility of deeper intervention changes the bargaining environment. Microsoft now has a strong incentive to offer incremental concessions early, before the CMA hardens its view of what meaningful competition restoration requires.

• Interoperability obligations are likely to be central.
• Licensing transparency could become a key remedy.
• Data portability and migration rules are probable candidates.
• Structural separation would be much harder, but not impossible to imagine in a long-running dispute.

Why timing is everything​

The CMA’s process may take months, and that creates a familiar regulatory tension: the market keeps moving while the case grinds on. By the time a final decision arrives, Copilot, Microsoft 365 packaging, and cloud migration practices may all have evolved again. Speed matters, because delayed remedies can end up solving yesterday’s competition problem.
Still, the regulator’s willingness to act before harm becomes entrenched is itself a statement. It suggests the UK now views digital competition as an area where waiting for market self-correction is too optimistic. That is a meaningful shift in regulatory philosophy, and Microsoft is one of the first companies to feel it in full.

---

Strengths and Opportunities​

The upside of this investigation is that it could finally clarify where the boundaries lie between legitimate integration and market foreclosure. A clear rulebook would help customers, competitors, and Microsoft itself by reducing uncertainty about what is acceptable in cloud licensing and AI bundling. It may also push the market toward better portability and more honest pricing.
It could also produce practical benefits well before the final legal outcome. Even the threat of SMS designation may encourage Microsoft and its rivals to improve switching tools, publish clearer terms, and reduce hidden lock-in mechanics. In that sense, the investigation itself may be part of the remedy.

• More transparency in cloud and software pricing.
• Better interoperability across providers.
• Lower switching friction for businesses.
• Stronger competitive pressure on hyperscalers.
• More room for UK challengers to grow.
• A clearer framework for AI bundling and productivity software.
• Potentially lower long-term procurement costs for enterprises.

Risks and Concerns​

The biggest risk is that the probe becomes too broad and too slow to deliver useful relief. If the CMA tries to regulate cloud, productivity software, and AI all at once without tightly defined objectives, it could create years of uncertainty with little immediate consumer benefit. That would be especially unfortunate in a fast-moving market where timing is crucial.
Another concern is remedy design. If rules are too vague, they will be gamed; if they are too prescriptive, they may ossify product innovation and create compliance burdens that smaller firms cannot bear. There is also the possibility that Microsoft simply adapts its packaging while preserving the economic effect of lock-in in new ways. That would not be a win for competition.

• A long investigation could delay practical relief.
• Overly broad remedies may be hard to enforce.
• Poorly designed rules could slow innovation.
• Microsoft could repackage products around the spirit of the rules.
• Compliance costs may fall unevenly on smaller rivals.
• Multi-cloud complexity may increase if rules are not technically precise.
• The CMA may struggle to keep pace with rapid AI product changes.

---

Looking Ahead​

The next phase will likely determine whether this is a symbolic probe or a genuinely market-shaping one. If the CMA launches the SMS investigation in the timeframe reported and frames it around software licensing and AI bundling, it will have effectively declared that the old cloud competition story is now inseparable from the modern productivity stack. That is a big statement, and it will carry implications far beyond Microsoft.
The broader market will also be watching for signals from Brussels and other regulators. The UK often acts as a test bed for digital competition policy, and if it succeeds in forcing clearer portability and fairer licensing, similar approaches could spread. If it fails, hyperscalers will take that as evidence that integration still outruns regulation.
What to watch next:

• Whether the CMA formally launches an SMS probe in May 2026 as reported.
• Whether Microsoft 365, Windows, and Copilot are all included in scope.
• Whether the CMA prioritizes licensing remedies over infrastructure remedies.
• Whether Microsoft offers further voluntary concessions before any formal designation.
• Whether rivals and UK cloud providers intensify their advocacy for stricter rules.

The most important thing to understand is that this is no longer a narrow dispute about cloud pricing. It is a test of whether the UK can regulate an ecosystem where software, cloud, and AI increasingly reinforce each other. If the CMA gets this right, it could make it easier for businesses to choose technology on merit rather than inertia. If it gets it wrong, Microsoft’s integrated model may become even harder to challenge than it is today.

---

Source: TechRadar Microsoft set to face another major UK probe - this time over cloud and software licensing