EU opens antitrust probe into the entire AI technology chain of the biggest tech corporations

The European Commission has launched a sweeping antitrust investigation into tech giants, targeting the entire AI technology chain, from computing infrastructure to end-user applications.

What the regulator is examining

The EU's antitrust chief said the regulator is studying possible distortions of competition across the whole structure of the AI market. This is a fundamental widening of scope compared with past cases: previously the European Commission dealt with individual products and practices, whereas now it is dealing with systemic market power as such.

The focus is on the vertical integration of the largest platforms. Amazon, Google, Microsoft and Meta* simultaneously control several critical layers of the market, which can potentially shut out independent players:

• Cloud infrastructure with priority access to scarce GPUs (Azure, GCP, AWS)
• Huge troves of training data: search queries, maps, email, messengers
• Strategic investments in leading AI labs (OpenAI, Anthropic, Mistral)
• AI assistants built directly into operating systems, browsers and office suites
• API access terms that can put third-party developers at a disadvantage

Why the investigation is opening now

Pressure on Big Tech in AI has built up gradually. The AI Act, which entered into force in 2024, set mandatory requirements for high-risk systems and foundation model providers. The Digital Markets Act (DMA) designated the six largest platforms as "gatekeepers" and required them to ensure interoperability with competitors. In practice, this means any restriction on compatibility can now be treated as a violation.

But the direct catalyst for the current investigation was the large investment alliances of recent years. Microsoft invested more than $13 billion in OpenAI and received exclusive rights to commercial hosting of GPT models via Azure. Amazon invested up to $4 billion in Anthropic. Google is both a major shareholder in Anthropic and its key cloud partner. This concentration of capital has already drawn regulators' attention: the FTC in the US and the CMA in the UK are running parallel reviews, and the European Commission is joining an emerging global consensus.

The semiconductor market is a separate problem. NVIDIA holds over 80% of the market for AI training accelerators, while the largest cloud platforms secure priority in supply queues. Independent startups get access to compute later and at a much higher price, which gives integrated giants a structural advantage as early as the model development stage.

"The entire AI technology chain is under our antitrust scrutiny: we are looking for possible distortions of competition at every level," a

European Commission spokesperson said.

What Big Tech risks

The potential sanctions are substantial. Under the DMA, "gatekeepers" face fines of up to 10% of global annual revenue for violations, and up to 20% for systematic violations. For Microsoft or Google, that means tens of billions of euros. Google's total EU fines for abuse of its position in search, Android and advertising have already exceeded €8 billion, a precedent that shows the regulator's readiness for tough decisions.

Structural remedies could hurt more than financial penalties. The European Commission has the right to demand open third-party access to cloud infrastructure, ban exclusive partnerships with AI labs, or require disclosure of API terms, up to forced divestiture of individual business units.

What this means

The regulatory logic is changing fundamentally, from piecemeal complaints about individual products to assessing structural market power as a system. For independent AI developers and European companies, forced opening of access to key resources (compute, training data, distribution channels) could become a turning point in the competition for the AI market of the next decade.

*Meta is designated an extremist organization and banned in Russia.