On 16 October 2026 the UK Competition and Markets Authority proposed designating Microsoft's business software ecosystem with Strategic Market Status (SMS) under the Digital Markets, Competition and Consumers Act 2024. Designation alone imposes no obligations. But it opens the door to the most consequential intervention in enterprise software in the UK since the cloud market investigation. The question for innovation policy is what comes through that door.
What the CMA actually proposed
The proposed decision covers five product areas: PC operating systems, productivity software suites, security software, server operating systems and relational database management systems. The CMA provisionally finds that Microsoft has substantial and entrenched market power and a position of strategic significance across this ecosystem. TokenPost's report says Microsoft had more than one million organisational customers in the UK in 2025, out of roughly 2.7 million UK businesses, and about 20 million monthly users.
The CMA case page records that the investigation opened on 14 May 2026 under Part 1 of the 2024 Act. Responses on the proposed decision, and on a separate roadmap of possible interventions, are due by 13 November 2026. The statutory deadline for the final decision notice is 13 February 2027.
The CMA stresses that SMS designation does not assume wrongdoing. The roadmap flags cloud software licensing, AI features such as Copilot, Teams and security bundling as areas for further work. Reporting on the proposal is explicit that these are possible subjects for scrutiny, not commitments to any particular measure.
The strongest case for the CMA
The argument for acting is serious and rests on evidence the CMA has already gathered. Its cloud services market investigation closed in July 2025. As the CMA's 31 March 2026 announcement describes, it found that Amazon and Microsoft hold significant market power in cloud and that Microsoft's licensing of key business software limits customer choice. Ordinary competition law is slow and case-by-case. A customer who already runs Windows Server, SQL Server and Microsoft 365 faces real switching costs. If licensing terms make running those products on a rival cloud more expensive, rivals compete at a structural disadvantage. A regulator that can act before the market tips further has a fair argument for doing so.
The bundling concern is also not hypothetical. When a dominant productivity suite adds an AI assistant, a collaboration app and security tooling to a single licence, rivals in each category must compete against a price of effectively zero. That can be efficient for customers. It can also foreclose competitors who might have built better products.
Where the proportionality risk sits
Designation is the easy part. The risk lies in how the follow-on conduct requirements are drafted. Three tests matter.
- Evidence of harm, not category membership. Bundling is often how software improves. A security feature built into an operating system raises the baseline protection of millions of small organisations that would never buy a standalone product. Requirements that prohibit integration by default would trade measurable security and usability gains for a speculative competition benefit.
- Licensing remedies that target the identified problem. The cloud finding concerns how Microsoft prices and licenses its software on rival clouds. A remedy aimed at that, such as non-discriminatory licensing terms, is proportionate. Broad design mandates for Teams or Copilot are not, unless the evidence shows they cause foreclosure.
- Room for AI to develop. Generative AI features in productivity software are still changing monthly. Locking in interoperability or choice-screen rules around today's Copilot risks regulating a product shape that may not exist in two years. Rivals' ability to compete in AI is a legitimate concern, as the CMA's launch notice acknowledges. But the remedy should wait for evidence of actual exclusion.
What the process gets right, and what to watch
To its credit, the CMA is using the regime in stages. It published the roadmap alongside the proposed decision, so Microsoft and its rivals can comment on possible interventions before any are imposed. The 2024 Act also requires that conduct requirements be proportionate and tied to an identified concern, and the CMA has said any intervention must meet the relevant legal tests. That sequencing is better than the alternative of designating first and improvising later.
There are still three things to watch. First, the CMA's own description of its approach as flexible and pragmatic is welcome, but flexibility must not become an open-ended licence to revisit the commercial terms of an entire product suite. Second, the interaction with the CMA's cloud commitments from Microsoft and Amazon on egress fees and interoperability matters: if voluntary steps already address part of the licensing problem, conduct requirements should fill only the remaining gap. Third, the UK is a mid-sized market. A heavy-handed regime could make it a place where Microsoft ships features later, or in cut-down form, which would cost UK businesses more than the intervention gains them.
Bottom line
Designating Microsoft is consistent with the evidence the CMA has published, and it is not an accusation. The substantive judgment comes after 13 February 2027, when the CMA must decide which of the roadmap's items are worth acting on. A narrow, licensing-focused first set of conduct requirements, monitored for results before anything touches product design or AI features, would deliver the competition benefit with minimal collateral damage to innovation. Respondents to the consultation should push the CMA to commit to that sequence in writing.