EU Asks Google for Details on AI Training
The European Commission opens a new chapter in its long-running scrutiny of Big Tech. This time, the spotlight lands on Google, its AI models, and the AI Summaries that appear in Google Search. Regulators want answers. They want clarity on how Google trains models, uses online content, and presents AI-generated text to millions of users across Europe.
Google says its models help users navigate information faster. EU officials say they need to ensure fairness, transparency, and compliance with the Digital Markets Act (DMA). This tension sets the stage for a high-stakes inquiry that touches everything from training data to competitive positioning.
What’s Happening & Why This Matters
The European Commission sends a formal request to Google for documents that outline how the company collects and uses online content to train its Gemini and other foundation models. Officials want to know what data Google pulls, whether publishers can opt out, and how copyright rules apply. According to the Commission, the request centres on transparency and compliance as required under the DMA.
Google states that it “works constructively with regulators” and follows legal frameworks in each market. The company maintains that AI training practices sit within fair-use norms and existing agreements with publishers.
AI Summaries in Google Search
Alongside training questions, the EU examines AI Summaries, the feature that displays machine-generated answers above search results. European publishers argue that these summaries reduce traffic to original sites. They claim that placement at the top of the results page compresses visibility and disrupts long-standing economic models in journalism.
The Commission asks Google to explain how AI Summaries rank information, what sources they rely on, and whether they treat all publishers fairly. Regulators want to know whether the summaries draw on content that publishers did not intend for training or summarisation.
Reaction and Signs
Newsrooms across Europe support the investigation. Several industry groups say Google must disclose how AI-generated text shapes audience behaviour. One media representative says, “We support innovation, but we need transparency. Visibility matters for democracy.”
Google responds by emphasising user benefit. The company points to tests showing that AI Summaries route users to deeper results and offer faster paths to relevant information. Critics argue that the summaries remove the need to click. This is the core friction between platforms and publishers.
TF Summary: What’s Next
The EU sets a clear tone. Regulators push for transparency, accountability, and clear data-use boundaries across every prominent AI platform. Google now faces a series of deadlines to share documents and provide technical explanations. These disclosures shape the next phase of AI governance across Europe. Publishers, lawmakers, and global competitors watch how the investigation unfolds.
MY FORECAST: Brussels augments its inquiries. The EU presses every AI provider for detailed disclosures. Expect new rules around training data, opt-out mechanisms, and the presentation of AI-generated text in search results. European publishers gain leverage as discussions around fair compensation and AI transparency speed up.
— Text-to-Speech (TTS) provided by gspeech