EU Compels Google to Share Search Data; Gemini Uses It to Personalize Generated Images

Google’s compliance timetable under the Digital Markets Act (DMA) now includes a concrete data‑sharing schedule that will reshape the European search ecosystem, according to the European Commission’s preliminary findings reported by The Register. The Brussels regulator has drafted a “to‑do list” that obliges the company to provide rivals with its core search‑related datasets—rankings, query logs, click‑through rates and view counts—on fair, reasonable and non‑discriminatory (FRAND) terms. By mandating access to the “search crown jewels,” the Commission aims to dismantle the data moat that has historically insulated Google Search from competition and to enable emerging search engines and AI‑driven chatbots to build services that can genuinely challenge the incumbent (The Register, 16 April 2026).

The policy shift is more than a procedural tweak; it is a direct response to the growing convergence of search and generative AI. Google’s own Gemini model, now in its Nano Banana 2 iteration, already leverages personal user data—Gmail, YouTube history, and Google Photos—to tailor image generation, as detailed by Engadget (16 April 2026). The “Personal Intelligence” feature allows Gemini to infer context from a user’s account without explicit prompts, producing images that reflect an individual’s preferences, labeled contacts, or frequently viewed content. This capability demonstrates how tightly Google’s AI products are intertwined with the very data the DMA seeks to open up to competitors. If rivals gain comparable access, they could replicate or even surpass Gemini’s personalized output, eroding Google’s differentiation advantage.

From a market‑structure perspective, the Commission’s move could accelerate the emergence of a multi‑player search landscape in Europe. Teresa Ribera, the EU’s executive vice‑president for Clean, Just and Competitive Transition, emphasized that “access to this data should not be restricted in ways that could harm competition” and warned that “small changes can quickly have a big impact” in fast‑moving markets (The Register). By forcing Google to share its data on FRAND terms, the EU is effectively lowering the barrier to entry for firms that can combine raw search signals with advanced AI models. Analysts have long noted that the value of search data lies not only in raw query volume but also in the nuanced ranking signals that inform relevance judgments. With those signals disclosed, new entrants could train proprietary models that rival Google’s performance, potentially reshaping ad‑tech revenue flows that currently flow predominantly to Google’s ad platform.

The timing of the DMA enforcement coincides with heightened scrutiny of Google’s data practices beyond competition law. The same week The Register reported that the EU’s public‑sector push is “angering Google,” reflecting broader concerns about privacy and the concentration of user‑generated content within a single ecosystem. Meanwhile, Engadget’s coverage of Gemini highlights a user‑centric narrative: the model’s ability to pull images from a user’s own Google Photos library and to cite sources transparently could set a new standard for AI explainability. If competitors adopt similar transparency tools, they may gain a regulatory edge, as EU policymakers have signaled a preference for AI systems that can disclose data provenance (The Register; Engadget). This regulatory climate may incentivize rivals to prioritize open‑source or interoperable AI stacks that can ingest the shared search data while complying with emerging EU AI governance frameworks.

In practice, the rollout will test both technical and legal boundaries. Google must design APIs that deliver granular search metrics without compromising user privacy—a challenge the Commission acknowledges when it notes that “data is a key input for online search and for developing new services, including AI” (The Register). Simultaneously, the company must preserve the integrity of its ad‑targeting algorithms, which rely on the same datasets now being handed to competitors. How Google balances these competing imperatives will likely dictate the speed at which the European market diversifies. If the data‑sharing mechanisms prove robust and privacy‑preserving, the DMA could catalyze a wave of AI‑enhanced search services that leverage both public query trends and individualized user signals, narrowing the gap between Google’s integrated ecosystem and the broader AI‑driven search frontier.