EU regulators identify gatekeepers' AI advantages in data and infrastructure access

A joint paper published by members of the High-Level Group on Artificial Intelligence outlined how the Digital Markets Act could address competitive imbalances in AI development, identifying two interconnected areas where gatekeepers may establish unmatchable advantages. The analysis preceded the European Commission's December 9, 2025 formal antitrust investigation into Google's use of publisher and YouTube content for AI purposes.

The paper, produced by the HLG subgroup on AI established under Article 40 of the Digital Markets Act, examined regulatory intersections across multiple frameworks including competition law, data protection rules, consumer protection legislation, and media regulations. According to the document, the subgroup convened to "discuss, provide advice, expertise and promote a consistent regulatory approach across the different regulatory instruments in relation to AI."

Members of the High-Level Group represent regulatory bodies including BEREC, EDPB, EDPS, ECN, and EBMS. The Commission invited these authorities to map potential contestability and fairness issues across the AI value chain, identifying areas where regulatory consistency proves essential and examining where different legal frameworks interact when addressing AI technologies.

The paper emphasized that gatekeepers enjoy "a strong presence at various levels of the AI value chain," particularly regarding general-purpose AI models increasingly integrated within gatekeeper ecosystems. This positioning allows designated platforms to influence multiple stages of AI development from infrastructure provision through end-user distribution.

Access to AI infrastructure and distribution channels

The HLG analysis identified AI services' strong reliance on cloud computing infrastructures "currently concentrated around very few players" as a primary contestability concern. According to the paper, gatekeepers rapidly integrate AI technologies into their services, creating gateways through which digital service providers must pass to reach end users.

"Such increasing level of integration may be a significant obstacle to third parties' ability to offer customised AI services to end users, and by the same token create disincentives for end users to seek third parties' offers," the document stated. The concentration creates incentives and opportunities for self-preferencing by gatekeepers while potentially locking in users through lack of effective interoperability with alternatives.

The paper specifically noted that when gatekeepers embed AI technologies into core platform services, "this may have implications also for content providers on the Internet in particular with regard to their traffic volume and visibility." This observation directly connects to concerns about AI-generated search features reducing publisher traffic.

Competition lawyer Thomas Höppner, a partner at GERADIN specializing in Digital Markets Act litigation, commented on the paper's significance in a December 2025 LinkedIn post. Höppner emphasized that the High-Level Group "rightly identifies two interconnected areas where the DMA can enhance contestability and fairness in the AI sector."

Höppner highlighted the paper's focus on access to AI infrastructure and distribution, noting the analysis of how increased integration "may be a significant obstacle to third parties' ability to offer customised AI services to end users." He quoted the document's finding that this integration "creates incentives and opportunities for self-preferencing by the gatekeepers, and may lock-in users due to a lack of effective interoperability with alternatives."

The lawyer particularly stressed the paper's examination of content provider implications. "Furthermore, when gatekeepers embed AI technologies into their core platform services, this may have implications also for content providers on the Internet in particular with regard to their traffic volume and visibility," Höppner wrote, emphasizing this point's connection to the Commission's new investigation.

Exclusive data access and partnerships create competitive advantages

The second interconnected area identified in the HLG paper concerned access to data for AI development. According to the analysis, "the ability to establish exclusive partnerships coupled with exclusive access to first-party data for use in Generative AI development may create an unmatchable advantage for gatekeepers compared to third parties."

The paper distinguished between different types of training data. While public data, pretraining datasets, and fine-tuning data remain important for AI development, "user interaction data appears key to ensure contestability in the AI sector." This category of data often consists of personal, first-party, proprietary information subject to various data protection legislation.

Gatekeepers do not make such data available to third parties, according to the document. "Network effects and feedback loops relating to data obtained by the gatekeepers from the use of these AI systems further exacerbate the data-driven advantages that gatekeepers already enjoy," the analysis concluded.

Höppner characterized this as addressing "access to data and content." He noted that "the ability to establish exclusive partnerships coupled with exclusive access to first-party data for use in Generative AI development may create an unmatchable advantage for gatekeepers." The lawyer emphasized that when gatekeepers embed AI technologies into core platform services, consequences extend to content providers regarding "their traffic volume and visibility."

"As painfully illustrated by Google's AI Overviews and AI Mode, the later point goes to the heart of the new competition investigation the EU opened this week," Höppner wrote. He connected the HLG's analytical framework directly to the Commission's December 9 enforcement action.

The lawyer further noted that "the need to scrutinize 'preferred [foreclosure] partnerships' was further underlined by this week's announcements about new partnerships by Google, Meta and OpenAI with some content providers, potentially to the exclusion of others." He expressed satisfaction that "the HLG appears to align with the concerns expressed by DG Comp."

Commission investigation targets publisher and YouTube content usage

The European Commission's December 9, 2025 press release announced formal proceedings examining whether Google breached Article 102 of the Treaty on the Functioning of the European Union and Article 54 of the European Economic Area Agreement. The investigation will assess whether Google imposed unfair terms on publishers and content creators while granting itself privileged access to content that places rival AI model developers at a disadvantage.

According to the Commission, Google may have used web publisher content to provide AI Overviews and AI Mode on search results pages "without appropriate compensation to publishers and without offering them the possibility to refuse such use of their content." The regulatory body noted that many publishers depend on Google Search for user traffic and "do not want to risk losing access to it."

AI Overviews shows AI-generated summaries above organic results, while AI Mode functions as a search tab answering queries in conversational style. The Commission will investigate "to what extent the generation of AI Overviews and AI Mode by Google is based on web publishers' content without appropriate compensation for that, and without the possibility for publishers to refuse without losing access to Google Search."

The investigation also examines YouTube content usage. Content creators uploading videos to the platform have "an obligation to grant Google permission to use their data for different purposes, including for training generative AI models," according to the press release. Google does not remunerate YouTube content creators for this use, nor does it allow them to upload content without granting such permission.

Meanwhile, rival AI model developers face barriers. "Rival developers of AI models are barred by YouTube policies from using YouTube content to train their own AI models," the Commission stated. This asymmetry grants Google exclusive access to training data while preventing competitors from accessing the same material.

Teresa Ribera, Executive Vice-President for Clean, Just and Competitive Transition, provided a statement emphasizing European values. "A free and democratic society depends on diverse media, open access to information, and a vibrant creative landscape," Ribera stated. "These values are central to who we are as Europeans. AI is bringing remarkable innovation and many benefits for people and businesses across Europe, but this progress cannot come at the expense of the principles at the heart of our societies."

Ribera explained the investigation's purpose: "This is why we are investigating whether Google may have imposed unfair terms and conditions on publishers and content creators, while placing rival AI models developers at a disadvantage, in breach of EU competition rules."

Google announces publisher partnerships and source preference features

One day after the Commission's announcement, Google published a blog post on December 10, 2025 titled "Supporting the web with new features and partnerships." Authors Robby Stein, VP of Product for Google Search, and Jafer Zaidi, VP of Global News Partnerships, outlined initiatives the company characterized as connecting people to the web and updating partnerships "for the AI era."

According to the blog post, Google launched Preferred Sources globally, a feature allowing users to customize Top Stories to see more from favorite outlets. "We're now launching this feature globally: in the coming days, it will be available for English-language users worldwide, and we'll roll it out to all supported languages early next year," the authors wrote.

The company reported that people selected "nearly 90,000 unique sources, from local blogs to global news outlets" as preferred sources. "When someone picks a preferred source, they click to that site twice as much on average," according to the post.

Google also announced "a new feature that highlights links from your news subscriptions, making it easier to spot content from sources you trust and helping you get more value from your subscriptions." The company stated it will "prioritize links from your subscribed publications, and show these links in a dedicated carousel" starting in the Gemini app, with AI Overviews and AI Mode to follow.

Regarding AI experiences, Google described increasing "the number of inline links in AI Mode, and updating the design of those links to make them more useful." The company added contextual introductions to embedded links explaining "why a link might be helpful to visit."

The blog post announced "a new commercial partnership program with a range of news publishers globally" including Der Spiegel, El País, Folha de S. Paulo, Infobae, Kompas, The Guardian, The Times of India, The Washington Examiner and The Washington Post. According to the announcement, these partnerships aim "to explore how AI can help drive more engaged audiences."

"As part of this new AI pilot program, we're working with publishers to experiment with new features in Google News," the authors wrote. Examples included "AI-powered article overviews on participating publications' Google News pages to give people more context before they click through" and audio briefings for those preferring listening. "These features will include clear attribution and link to articles," the post stated.

Google also announced partnerships with Estadão, Antara, Yonhap and The Associated Press "to include real-time information to enhance results in the Gemini app."

Regulatory framework interplay across multiple legal instruments

The HLG paper mapped extensive touchpoints between the Digital Markets Act and other regulatory frameworks when applied to AI technologies. The analysis identified areas where multiple legal instruments address similar practices from complementary angles.

Regarding data portability, the paper noted that "the users' right to port data can also be found in distinct instruments" including the Data Act's contractual obligations for cloud services providers, the GDPR's data portability requirements, and the DMA's user portability rights. The document suggested benefits from "considering the interplay across these provisions and striving for alignment."

The paper examined dark patterns across the DSA, AI Act, UCPD, GDPR and DMA. These frameworks aim to prevent manipulative design "including in AI interfaces e.g., undisclosed commercial intent in AI-generated search results," according to the analysis. The Commission suggested a need to "ensure consistency by agreeing on 'clear and comprehensible' disclosure standards that would be adequate across the various relevant provisions in these regulations."

Transparency requirements also intersect across multiple frameworks. The paper identified connections between the AVMSD's restrictions on targeted ads for minors, the DMA's obligations on ad transparency for gatekeepers, and DSA provisions on recommender system transparency. Additionally, "the European Media Freedom Act (EMFA)'s editorial independence provisions have an interplay with the AI Act's transparency obligations for high-risk AI systems," the document noted.

The analysis highlighted algorithmic explainability rights appearing in different forms across regulations. The GDPR explicitly references individual rights to explanation in automated decision-making contexts, which "might operate complementarily to the right to an explanation of individual decision-making provided under Article 86 of the AI Act," according to the paper.

Competition lawyer Thomas Höppner serves as a partner at GERADIN and focuses on Digital Markets Act litigation. His December LinkedIn commentary on the HLG paper attracted engagement from other regulatory observers. In response to questions about Google's new AI partnerships, Höppner directed readers to specific URLs, noting "On Google's new AI partnerships see here" followed by the Google blog link, and "on the EC's new probe into GenAI see here" with the Commission press release URL.

Another commenter, identified as a competition lawyer, technology specialist and travel/hospitality professional, responded to Höppner's post stating agreement with the need for legal frameworks. Rosalia Anna D'Agostino wrote that "we need legal frameworks, asap, backed up with other settings to foster plurality, guarantee diversity and fair competition."

Cross-regulatory cooperation identified as essential mechanism

The HLG paper concluded that "cross-regulatory cooperation between competent authorities would help facilitate clarity and coherence when it comes to the interplay between the multiple frameworks of legislation applicable to the AI value chain." The document emphasized particular relevance "where multiple frameworks apply to the same practices or operations, such as the development and deployment of AI systems by gatekeepers."

According to the analysis, coordination between regulatory bodies dealing with AI technologies constitutes "a crucial aspect of ensuring consistent enforcement." The paper stated that "all relevant regulatory frameworks need to be interpreted coherently to avoid over- and underenforcement."

The document called for developing "a common understanding of AI related issues" with enforcement practices reflecting such understanding. "Overall, coordination between different regulatory bodies dealing with AI is essential to ensuring that AI is developed and deployed in a way that is consistent with EU values and principles," the paper concluded.

The analysis referenced the Court of Justice of the European Union's recent Meta Platforms Inc and Others v Bundeskartellamt judgment as highlighting cooperation requirements. That ruling emphasized "that authorities at both national and EU-level are bound by the principle of sincere cooperation when enforcing EU law," according to the HLG paper.

"Cooperation and coordination in this regard should be made more effective between the competent authorities," the document stated. The paper suggested that the HLG's mandate under the DMA "covers this and allows enhanced coordination and cooperation between the relevant regulatory bodies within the forum of the HLG as provided for in the DMA."

Regarding small and medium-sized enterprises, the paper highlighted their importance "both in order to enable them to enjoy the benefits of the DMA, as they are often the main beneficiaries of DMA obligations, and as the real economic actors that stand to benefit from the impact of the DMA."

The analysis noted that SMEs require legal certainty and should not face "overly burdensome barriers to compliance with AI-related regulation." Specific opportunities mentioned included how "interoperability obligations can open opportunities in markets that used to be controlled by gatekeepers" and how "data related obligations can grant SMEs access to this critical input in the development of AI services."

The Commission's DMA team has "already started outreach to several SMEs that provide AI services and has received positive feedback from these interactions," according to the paper. These discussions provided "a high-level description of the DMA and its obligations to gatekeepers" while emphasizing that SMEs "will benefit from coherent enforcement of different regulations, that can result from the inter-agency cooperation discussed above, as this will lead to legal certainty and easier compliance."

Investigation timeline and procedural framework

The Commission's December 9 press release specified that there is "no legal deadline for bringing an antitrust investigation to an end." According to the announcement, "the duration of an antitrust investigation depends on a number of factors, including the complexity of the case, the extent to which the companies concerned cooperate with the Commission and the parties' exercise of the rights of defence."

The Commission informed Google and member state competition authorities that it opened proceedings in this case. According to Article 11(6) of Regulation 1/2003, "the opening of proceedings by the Commission relieves the competition authorities of the Member States of their competence to also apply EU competition rules to the practices concerned."

Article 16(1) further provides "that national courts must avoid adopting decisions which would conflict with a decision contemplated by the Commission in proceedings it has initiated," the press release noted. The investigation case number is AT.40983, with information available on the Commission's competition website public case register.

The press release emphasized that "the opening of a formal investigation does not prejudge its outcome." The Commission stated it "will now carry out its in-depth investigation as a matter of priority."

If the investigation proves Google's practices violated competition rules, potential consequences could include significant fines and behavioral or structural remedies. Article 102 of the Treaty on the Functioning of the European Union prohibits abuse of dominant position, with implementation defined in Regulation 1/2003.

The European Commission has pursued multiple antitrust cases against Google over the past decade, addressing search practices, Android operating system requirements, and advertising technology. The Court of Justice upheld a €2.4 billion fine for Google Shopping practices in September 2024.

Brussels imposed a €2.95 billion fine on September 5, 2025 for advertising technology violations, finding Google favored its own services over competitors. Publishers have filed antitrust complaints targeting AI Overviews features, alleging traffic declines when AI summaries appear in results.

Marketing implications for advertising professionals

The investigation's outcomes could reshape competitive dynamics in digital advertising and search marketing. Publishers reporting substantial traffic decreases when AI Overviews appear face pressure on advertising revenue models that depend on user visits to websites.

Research has documented click-through rate declines ranging from 34.5 percent to 54.6 percent when AI summaries display in search results. These metrics directly impact publisher monetization through reduced advertising inventory and audience engagement opportunities.

Marketing professionals working with publisher partnerships face uncertainty about long-term platform strategies. The Digital Markets Act has already prompted significant changes to Google Search functionality in European markets, with the company implementing over 20 modifications to comply with self-preferencing prohibitions.

Google rejected EU demands for advertising technology asset sales on November 13, 2025, instead proposing product modifications. The company faces parallel proceedings in the United States where federal courts have ruled on monopolization claims in both search and advertising technology markets.

The HLG paper's identification of AI-related competitive concerns suggests regulators will scrutinize how platforms leverage market positions across the AI value chain. Marketing professionals should anticipate potential changes to how AI-powered features integrate with search results, advertising placements, and content distribution mechanisms.

The investigation examines fundamental questions about platform obligations when using third-party content for AI development. Outcomes could establish precedents affecting how technology companies compensate content creators and whether competitors receive access to training data currently controlled by dominant platforms.

Timeline

• 2010: European Commission launches antitrust investigation into Google's search practices following complaints
• 2017: Commission imposes €2.4 billion fine for Google Shopping violations
• May 2, 2023: Digital Markets Act designates six gatekeepers including Google
• March 2024: DMA obligations become legally binding for designated gatekeepers
• April 10, 2025: High-Level Group subgroup on AI meets to discuss regulatory interplay
• June 30, 2025: Independent Publishers Alliance files antitrust complaint targeting AI Overviews
• September 5, 2025: Commission imposes €2.95 billion fine for advertising technology abuse
• November 13, 2025: Google rejects structural remedies in ad tech case response
• December 9, 2025: European Commission opens formal investigation into Google's AI content practices
• December 10, 2025: Google announces Preferred Sources global launch and publisher partnership program

Summary

Who: The European Commission launched an antitrust investigation into Google, the multinational technology company operating dominant search, video sharing, and digital advertising platforms. The High-Level Group on Artificial Intelligence, comprising representatives from BEREC, EDPB, EDPS, ECN, and EBMS, produced analysis examining Digital Markets Act application to AI competition. Competition lawyer Thomas Höppner provided expert commentary connecting the HLG paper to enforcement actions.

What: Brussels regulators opened formal proceedings examining whether Google violated EU competition rules through AI content usage practices. The HLG joint paper identified two interconnected areas where gatekeepers may establish competitive advantages in AI development: access to infrastructure and distribution channels, and exclusive access to first-party data through partnerships. The analysis mapped regulatory framework intersections across competition law, data protection, consumer protection, and media regulations as they apply to AI technologies.

When: The Commission announced the investigation on December 9, 2025. The HLG subgroup on AI convened on April 10, 2025 to discuss regulatory interplay issues. Google announced publisher partnerships and new features on December 10, 2025, one day after the investigation became public.

Where: The investigation proceeds under European Union jurisdiction applying Article 102 of the Treaty on the Functioning of the European Union and Article 54 of the European Economic Area Agreement. The case affects Google's operations across European markets and establishes precedents for how competition rules apply to AI development using content from web publishers and YouTube creators.

Why: Regulators identified concerns that gatekeepers leverage dominant positions across multiple stages of the AI value chain to create unmatchable advantages. The HLG paper emphasized that increased integration of AI technologies into core platform services may obstruct third parties' ability to offer customized AI services while creating self-preferencing opportunities. Access to exclusive first-party data combined with network effects and feedback loops exacerbates competitive imbalances. The investigation tests whether platforms can develop AI capabilities while respecting competition rules protecting content creators and rival developers.