The European Commission announces an antitrust investigation targeting Google, suspected of using content from YouTube publishers and creators to train its AI models and power AI Overviews and AI Mode without adequate compensation or the ability to object.
Brussels is investigating whether Google is exploiting its position in search and video to capture massive volumes of data, while locking out access for competing AI developers. In both cases, the Commission questions the existence of contractual conditions deemed unbalanced for publishers, creators and players in the AI market.
Publishers’ dependence on Google traffic, a central point
With AI Overviews and AI Mode, results pages now include generative summaries constructed from web content. Brussels is examining the absence of a remuneration mechanism and the difficulty for publishers to object to the use of their content without risking a loss of visibility in Google Search. While a major part of media traffic comes from the Google engine, the margin of refusal remains limited.
The investigation must establish whether this dependence creates a crowding out effect, at the same time that the responses generated reduce the incentive for users to click through to the source sites.
YouTube, a data reservoir reserved for a single actor
The other part concerns YouTube videos, so creators cannot publish without authorizing Google to use their content to train its models, and are not paid for this use. At the same time, the platform’s policies prohibit competing players from exploiting these videos to train their own models.
This locking of access to a database unique in its size and quality constitutes one of the most sensitive points of the file, because it risks creating a structural advantage difficult to fill for European AI companies, while the models require massive and varied volumes to remain competitive.
A systemic survey for the European AI ecosystem
The file comes at a time when Europe is trying to define a balanced framework between platforms, creators and publishers. The question of the value of content in the training of AI models, until now largely absent from competition law, now enters the field of analysis of the authorities.
Teresa Ribera recalls the political background of the issue.
“A free and democratic society depends on diverse media, open access to information and a vibrant creative landscape. These values are at the heart of who we are as Europeans. AI is bringing remarkable innovation and many benefits to people and businesses across Europe, but these advances cannot come at the expense of the principles that are at the heart of our societies. This is why we are examining whether Google may have imposed unfair conditions on publishers and content creators, while disadvantaging developers of competing AI models, in violation of EU competition rules. »
The investigation could become a structuring precedent for access to data in generative AI, in particular on transparency and remuneration obligations.
Google faces major regulatory risk
The Commission does not prejudge the result, but the scope of the case is reminiscent of the Google Shopping or Android precedents, with risks ranging from significant fines to structural injunctions on the use of content. The conclusions could also pave the way for an enforceable right of opt-out, licensing mechanisms or reinforced regulation of the conditions imposed by YouTube.
