news42.hu / Articles / Analysis

EU DMA and AI Search: What Changes for News Discovery

The Digital Markets Act forces Google to share search data with rivals, altering how readers find and verify news in an AI-driven era.

By Zoltán Varga2026-11-054 min read

What the DMA Requires of Search

The European Commission enforces the Digital Markets Act to curb the dominance of major technology gatekeepers. Under these rules, Google must provide access to its core search services for competitors. This mandate targets the structural advantage held by the incumbent search engine. The goal is to create a more level playing field for innovation and consumer choice.

Specifically, the regulation compels Google to share data used for search optimization. This data must be anonymized before transfer to ensure user privacy remains protected. The requirement takes effect in January 2027. Google faces a deadline to establish the technical infrastructure for this exchange. Failure to comply results in significant financial penalties.

The scope extends beyond simple data transfer. The act addresses the pre-installation of software on mobile devices. Google must allow users to choose default search engines and browsers without obstruction. This change dismantles the automatic routing of traffic that previously favored the incumbent. The regulation treats search as a critical utility rather than a proprietary feature.

Historical context shows a pattern of enforcement. In July 2018, the European Commission fined Google €4.34 billion for illegal practices regarding Android devices. That ruling focused on forcing manufacturers to preinstall Chrome and Google Search. The current DMA builds upon these precedents to address the rise of artificial intelligence. The legal framework evolves to cover new technological bottlenecks.

Data Sharing with Rivals and AI

Google must make its search optimization data available to rival AI chatbots. Companies like OpenAI and Anthropic will gain access to this information. They can use it to train their own models and improve response quality. The data is shared on a commercial basis with a defined fee structure. This creates a new revenue stream for the search giant while fueling competition.

The transfer includes anonymized signals that influence search rankings. These signals help rival models understand which sources users trust. The process does not reveal individual user identities. It focuses on aggregate patterns of behavior and relevance. This distinction is crucial for maintaining privacy standards while enabling interoperability.

Google retains the right to assess security risks before granting access. The company can vet potential partners for safety and compliance issues. This precaution aims to prevent misuse of the shared data. It adds a layer of complexity to the implementation timeline. The vetting process may delay full access for some competitors.

The move fundamentally changes how AI models learn about the web. Previously, these systems relied heavily on public crawling. Now they receive direct inputs from the dominant search engine. This access accelerates the development of alternative search tools. It reduces the barrier to entry for new market players. The ecosystem becomes more fragmented but potentially more diverse.

Effects on Referral Traffic

News publishers face uncertainty regarding future traffic sources. Traditional click-through rates from search results may decline. Users might read summaries directly within AI interfaces. This phenomenon is known as zero-click searching. The value of a link shifts from driving visits to establishing authority.

The correlation between high search rankings and AI citations is not guaranteed. Studies indicate that top positions in Google do not always translate to mentions in AI responses. Different models prioritize different signals. Some favor recent content, while others value historical authority. The logic varies by platform and specific query type.

Zero-click interactions can reduce overall referral volume. A user receives an answer without visiting the original website. Publishers lose the opportunity to engage readers deeply. Brand recognition may increase, but direct traffic could drop. This dynamic challenges existing business models for digital news.

Adaptation requires a focus on entity representation. Publishers must ensure their content is structured for machine interpretation. Technical SEO remains relevant but serves a different purpose. The goal is to be recognized as a credible source by various algorithms. Success depends on visibility across multiple AI systems, not just one search engine.

New Discovery Paths for Digests

AI-assisted digests will emerge as a primary discovery method. Readers will ask questions and receive synthesized answers. These answers pull from multiple sources simultaneously. The format favors concise summaries over long-form articles. News organizations must adjust their content strategies accordingly.

The rise of voice assistants expands these discovery paths. Eleven Android functions will open to rival AI assistants by July 2027. Voice commands will trigger searches across different providers. Users will hear news updates without typing queries. This shift favors audio-friendly content and clear, factual statements.

Curated feeds may integrate directly into operating systems. The boundary between the OS and the search engine blurs. Users access information through native interfaces. This integration reduces the need to visit external websites. It places a premium on being the preferred source for the system.

Readers must develop new verification habits. AI summaries can hallucinate or misattribute facts. Cross-referencing becomes essential for accuracy. Trust shifts from the platform to the underlying source. Media literacy skills become critical for navigating this environment. The responsibility for fact-checking moves closer to the end user.

Open Questions and Future Risks

The long-term impact on news diversity remains unclear. Concentration of power could shift from search engines to AI developers. A few large models might dominate information flow. This scenario poses risks for pluralism and viewpoint variety. Regulatory oversight will need to monitor these emerging concentrations.

Geopolitical tensions influence the implementation of these rules. American companies hold a significant advantage in model development. European entities struggle to compete without similar resources. The DMA opens the door, but market dynamics determine who walks through. Local language support varies widely across models.

Measurement of AI visibility lacks standardization. There is no universal metric for citation frequency. Publishers cannot easily track their performance in AI responses. This opacity hinders strategic planning. The industry needs transparent tools to assess reach and impact.

The timeline for full implementation involves significant delays. Security vetting and technical integration take time. The market will evolve slowly over several years. Current strategies may become obsolete before the changes fully take effect. Continuous adaptation is necessary for survival in this shifting landscape.

AI-assisted: drafted with a language model from the author's research notes, then edited and checked before publication. Author: Zoltán Varga, AI visibility researcher and AI ecosystem architect, Budapest.