When regulators become a player: antitrust, data law, and AI rules reshaping SaaS
In October 2023, Microsoft started selling Teams separately from Microsoft 365 in Europe. In April 2024 it did the same worldwide. Microsoft had bundled Teams into its office suite since 2017, and it did not unbundle it for strategic reasons. Slack (owned by Salesforce since 2021) filed an EU antitrust complaint in 2020, and the European Commission opened a formal investigation in 2023. In September 2025 the Commission made Microsoft's commitments legally binding: suites sold without Teams at a meaningful price difference, better interoperability for rival tools, and easier data portability, for up to ten years.
That case was brought under classic EU competition law (Article 102 of the EU treaties, on abuse of dominance). It was not brought under the newer Digital Markets Act. Both tools matter, and they work differently. Either way, rivals like Slack, Zoom, and Google Chat now compete on a more even footing in Europe. That is the core idea of this lesson: regulators are not referees standing outside the SaaS (Software as a Service) market. They redraw who competes with whom, and on what terms.
Regulators as a fifth force
Competitive analysis usually maps incumbents, challengers, suppliers, and distributors. In SaaS you need a fifth category: regulators, who can change market structure on their own.
The bodies that matter most:
- The European Commission, which enforces EU competition law, the DMA, and GDPRGDPREU regulation governing how organizations collect, store and use personal data, with fines tied to global revenue for breaches.View full definition → (the General Data Protection Regulation, the EU's 2018 privacy law). Day-to-day GDPR enforcement sits with national authorities.
- The US Federal Trade Commission (FTC) and Department of Justice (DOJ), which pursue antitrust cases under existing US law. There is still no US federal equivalent to the DMA.
- National data protection authorities. Ireland's Data Protection Commission leads on most US Big Tech because their EU headquarters are in Dublin. Germany has federal and state-level regulators.
A new competitor has to build a product, raise capital, and win customers. A regulator only has to write a rule or close a case with conditions, and incumbents restructure around it. Incumbents plan differently once they know this.
Case 1: The DMA and gatekeeper obligations
The DMA applies fully since March 2024. It designates large platforms as "gatekeepers" (companies that control access between businesses and users) and imposes obligations on them up front, without a years-long antitrust investigation first. The gatekeepers are Alphabet, Amazon, Apple, Booking, ByteDance, Meta, and Microsoft. For Microsoft, the designated services are Windows and LinkedIn, not Teams.
Obligations relevant to SaaS include:
- Interoperability: competing services must be able to connect to a gatekeeper's core platforms in specified ways.
- A ban on self-preferencing, so gatekeepers cannot rank their own services above rivals in search results or app stores.
- Data portability for business users and end users.
- Limits on tying, so users cannot be forced to sign up for one core service to use another.
The DMA has teeth. In April 2025 the Commission fined Apple €500 million over anti-steering rules in the App Store and Meta €200 million over its "pay or consent" ad model. For a SaaS company that sells through app stores, anti-steering rules directly affect margins. They decide whether you can point customers to cheaper web checkout and avoid platform commissions.
The gatekeeper obligations are listed on the European Commission's DMA page.
Case 2: GDPR and the rise of European cloud vendors
GDPR, in force since 2018, sets strict rules on how companies handle EU residents' personal data: consent, data minimization, and transfers outside the EU.
The transfer rules became a competitive weapon. In the 2020 *Schrems II* ruling, the EU Court of Justice struck down the EU-US Privacy Shield and found that US surveillance law conflicted with EU privacy rights. For three years, European companies using AWS, Azure, or Google Cloud for personal data faced real legal risk.
The EU-US Data Privacy Framework restored a legal basis for transfers in July 2023. The EU General Court upheld it in September 2025, though an appeal and further challenges remain possible. Many buyers treat it as fragile, since its two predecessors were both struck down.
The result is a durable "sovereign cloud" category. OVHcloud (France), Deutsche Telekom's T-Systems, and EU-localized offerings from the hyperscalers (such as the AWS European Sovereign Cloud, run from Germany) all sell data residencydata residencyThe requirement that data is physically stored and processed in a specific country or region, often driven by law or contract.View full definition → and EU legal jurisdiction as core features. France and Germany push public-sector procurement toward these options. France's SecNumCloud certification effectively excludes providers subject to US law unless they operate through European joint ventures.
This mechanism works differently from Case 1. Regulation did not break up a product here. It made geography and jurisdiction into product features.
How this changes power dynamics along the value chain
| Player | Effect of EU competition, DMA and GDPR-style regulation |
|---|---|
| Large US hyperscalers and suites (AWS, Azure, GCP, Microsoft 365) | Lose some bundling and data-transfer advantages; must fund compliance infrastructure (costly, but also a moatmoatA lasting edge over competitors: a resource, capability or position they cannot easily replicate, letting a firm earn above-average returns over time.View full definition → against smaller entrants) |
| Challengers (Slack, Zoom, regional SaaS vendors) | Gain forced access, interoperability and fairer pricing, at least in the EU |
| European SaaS/cloud vendors | Gain a differentiator (data sovereignty) that used to carry little commercial weight |
| Enterprise customers | Gain portability, more vendor choice, lower switching costs |
| Regulators themselves | Become an ongoing variable in strategy, not a one-off event |
Look at the compliance cost line. GDPR, DMA, or AI Act compliance is expensive, and large incumbents absorb it more easily than a 50-person startup. Regulation aimed at curbing incumbent power can raise barriers for the challengers it means to help. That tension is real.
The next frontier: the EU AI Act
The EU AI Act entered into force in August 2024. It classifies AI systems by risk. Bans on prohibited practices applied from February 2025. Obligations for general-purpose AI models (documentation, transparency, copyright policies) applied from August 2025. Stricter rules for "high-risk" uses such as hiring or credit scoring were due in August 2026. In November 2025 the Commission proposed a "Digital Omnibus" package that would push those high-risk deadlines back, possibly to late 2027. Check the current status before relying on any date.
For SaaS vendors embeddingembeddingAn embedding is a numerical vector that represents data (text, images, or items) in a way that captures meaning, so similar items sit close together in space.View full definition → AI features, which by now is nearly all of them:
- Foundation model providers (OpenAI, Anthropic, Google, Mistral) carry the general-purpose model obligations. Many signed the Commission's voluntary GPAI Code of Practice in 2025. Meta declined to sign. SaaS companies building on these models inherit questions from their own customers about it.
- An HR SaaS tool that screens candidates falls into the high-risk category, with conformity assessments and human-oversight requirements.
- Compliance costs again scale better for large players.
- The sovereign pattern may repeat. Mistral AI markets EU-based models as a differentiator. Aleph Alpha shifted from building frontier models toward enterprise and government AI tooling.
Watch 2026 and 2027 closely. The GDPR pattern is playing out again in a new layer of the stack, and the final timelines are still moving.
Knowledge check
1. What is the core conceptual shift the Microsoft Teams unbundling illustrates for competitive analysis in SaaS?
2. Why does the lesson describe regulatory power as a 'different kind of power' compared to a new competitor entering the market?
3. A US-based SaaS company primarily serving European customers is trying to determine which regulatory body most directly shapes its data privacy obligations. Based on the lesson, which of the following is most likely to have direct oversight?
4. Select ALL correct answers about why regulators are considered a 'fifth force' in SaaS competitive analysis.
Select all the correct answers.
5. Select ALL correct answers describing the regulatory bodies identified in the lesson as currently most influential in shaping global SaaS markets.
Select all the correct answers.
Reading the regulatory chessboard as a strategic skill
When you assess a SaaS company's competitive position, add these questions to the usual product and market ones:
- Is this company a gatekeeper, dominant in an antitrust sense, or competing against one? That status predicts future obligations or advantages.
- How exposed is its data architecture to transfer risk if the Data Privacy Framework falls? This affects EU enterprise sales cycles directly.
- Do its AI features fall into a high-risk category under the AI Act?
- Is regulation creating a moat here or removing one?
The European Commission's competition case search lists ongoing antitrust and DMA investigations against named companies.
Key Takeaways
- Regulators reshape SaaS competition, not just police it. Microsoft's Teams unbundling came from an EU antitrust case triggered by Slack, and was locked in by binding commitments in 2025.
- The DMA adds up-front rules for seven gatekeepers, and fines against Apple and Meta in 2025 show it is enforced.
- GDPR transfer rules, especially after *Schrems II*, turned legal jurisdiction into a product feature and built a lasting sovereign cloud market. The 2023 Data Privacy Framework eased the risk without removing it.
- Compliance costs cut both ways: they curb incumbent power and also favor large players who can absorb them.
- The EU AI Act is repeating the pattern in the AI layer, with model obligations already live and high-risk deadlines possibly delayed to 2027.