Competition regulation

AI and competition regulation

AI changes competition in two ways at once: it is a strategic capability that firms compete to develop, and it is also a market mechanism that can reshape behaviour. On the structural side, AI can reinforce concentration when a small number of firms control the inputs that matter most: compute, proprietary datasets, and the distribution channels where AI is embedded (operating systems, app stores, browsers, search, and major platforms). When those same firms integrate AI assistants and model features deeply into their ecosystems, competition can shift from ‘best model wins’ to ‘who controls the default interface for discovery, payments, identity, and attention’.

On the conduct side, AI-driven systems raise newer antitrust questions. Pricing algorithms, ad-auction optimisation, and recommender systems can produce market outcomes that look coordinated even without explicit human collusion. Regulators increasingly ask whether algorithmic optimisation can facilitate tacit coordination (firms independently converging on less competitive outcomes) or create discriminatory outcomes that are hard to detect because decisions are distributed across models and feedback loops. In practice, this is where competition regulation touches other governance domains: if a model systematically steers users toward the platform’s own services, that can look like a competition issue; if it does so in opaque ways, it becomes an accountability and transparency issue too.

AI and enforcement of competition regulation

Competition authorities increasingly rely on data analysis, pattern detection, and market monitoring to understand complex digital ecosystems. But enforcement becomes harder when key competitive levers are ‘inside the black box’: ranking systems, ad auctions, recommender systems, and model-integrated distribution channels that are constantly updated. When the rules of competition are effectively encoded in systems that outsiders cannot inspect, traditional antitrust’s case-by-case model can struggle to keep up, especially because markets may tip before remedies arrive.

That dynamic helps explain why several regulators are complementing classic enforcement with ex ante approaches (preventive and proactive measures) for the most powerful platforms. The aim is to reduce the risk that a dominant platform’s design choices become an irreversible market structure before authorities can intervene.

Learn more on AI Governance.

Did you know that competition in digital markets is no longer shaped only by price, but also by who controls visibility, access, and choice online? In today’s platform-based economy, power often lies in the ability to determine what users see first, which services come pre-installed, how easily rivals can reach customers, and whether people can switch without losing data, convenience, or social connections.

Competition regulation has become a key part of digital governance, concerned not only with keeping markets fair but also with ensuring that the architecture of the digital environment remains open and contestable.

 

Competition regulation

Competition regulation (often called antitrust) is the set of laws, enforcement tools, and remedies designed to keep markets open, contestable, and fair, primarily when one or a few firms can act as bottlenecks between businesses and users. In practice, it targets three recurring problems: cartels (illegal coordination), abuse of market power (such as self-preferencing, tying, exclusive dealing, or discriminatory access), and mergers that may reduce competition by removing future rivals. In digital markets, these concerns intensify because scale can become self-reinforcing: network effects make services more valuable as more people use them; data advantages can improve products and ads; and ecosystem design (e.g. single sign-on, app stores, default settings, bundled services) can mean lock-in or raise switching costs for users and business customers.

Concrete cases show what ‘abuse’ can mean in digital governance terms. The EU Commission’s decision in the Google Shopping case treated self-preferencing as a competition problem: a dominant search service allegedly gave its own comparison-shopping service privileged placement while demoting rivals, shaping what users see and what businesses can reach. The Android decision similarly focused on distribution power: practices around pre-installation, tying, and incentives were framed as limiting manufacturers’ and users’ ability to choose competing search and browser services. These examples matter for digital governance because they show that ‘competition’ is not only about prices; it is also about visibility, default pathways, access to interfaces, and the rules embedded in digital infrastructure.

 

How is competition regulation conducted?

In the classic model, authorities investigate and sanction harmful conduct after it occurs (ex post): cartel cases, dominance abuses, and merger control. Remedies used to focus on stopping a practice or imposing behavioural commitments. In the digital sphere, however, remedies increasingly target the infrastructure of competition: access to key interfaces, data portability, interoperability, and limits on self-preferencing. The policy logic is simple: when a platform becomes a bottleneck, competition may depend less on abstract ‘innovation’ and more on whether rivals can actually reach users on fair terms, and whether users can switch without losing functionality, data, or social connections.

 

Legal instruments and policy initiatives: the EU’s DMA, adjacent DSA rules, the UK’s Digital Markets, Competition and Consumers Act 2024, and some international examples

The EU – the DMA and DSA: In Europe, the Digital Markets Act (DMA) is the flagship ex ante instrument targeting gatekeepers, adding a rule-based layer to traditional EU competition enforcement. It reflects the view that some digital markets need upfront obligations to stop bottlenecks and unfair practices from hardening into durable dominance. The Digital Services Act (DSA) is not a competition law in the strict sense, but it sits next to this framework by imposing transparency, accountability, and recommender-system obligations on online intermediaries, which often affect how power is exercised in digital markets. Together, the DMA and DSA illustrate how the EU increasingly governs platform power not only through classic antitrust investigations, but also through operational rules on how large digital infrastructures must behave.

 

Two DMA themes show how competition regulation now overlaps with digital and AI governance. First, interoperability: Article 6(7) is designed to ensure that third parties can connect to key operating system features effectively and free of charge, rather than being disadvantaged by the platform owner’s control over defaults and core functionalities. Second, data access and contestability: Article 6(11) addresses access to certain search-related data on fair terms, reflecting a broader idea that some markets cannot become genuinely contestable unless rivals can access key inputs under safeguards. In this sense, competition law in digital markets increasingly reaches inside technical architectures, where default settings, APIs, ranking systems, and access conditions can shape competitive outcomes.

 

UK: The UK has pursued a similar direction through the Digital Markets, Competition and Consumers Act 2024, which establishes a forward-looking digital markets regime, enforced by the CMA’s Digital Markets Unit, for firms designated as Strategic Market Status holders. That regime took effect on 1 January 2025 and is designed to work alongside, rather than replace, the UK’s existing competition framework. Like the DMA, it signals a move away from relying solely on slow, after-the-fact antitrust cases and toward earlier intervention in markets where platform power can rapidly become entrenched.

 

The antitrust crash on Big Tech: Across jurisdictions, competition and digital-governance rules are now being brought into day-to-day enforcement in increasingly similar ways: regulators are moving from after-the-fact antitrust cases to ex ante obligations, faster investigations, and more operational remedies touching interoperability, app stores, search, advertising, and data access. In the EU, the DMA is already being operationalised through interoperability proceedings and specification measures, including requirements aimed at ensuring ‘free and effective’ interoperability on Apple’s iOS and iPadOS under Article 6(7). In the UK, the parallel shift is the DMCC Act regime now administered by the CMA. This broader convergence suggests that platform governance is becoming less about isolated headline cases and more about continuous supervision of how digital gatekeepers structure access to markets.

 

Japan and the US: Globally, several of the most consequential platform-focused moves follow this same logic. In Japan, the Mobile Software Competition Act creates ex ante rules for designated smartphone software providers in areas such as mobile operating systems, app stores, browsers, and search engines, making it one of Asia’s clearest equivalents to the EU’s gatekeeper model. In the US, the approach remains more litigation-driven, but it is still generating structural consequences: in the DOJ’s search monopolisation case against Google, the court ordered remedies including limits on exclusive distribution agreements and obligations to make some search-related data available to rivals. Together, Japan and the US exhibit two distinct regulatory styles: one rule-based and preventive, the other case-based and remedial, converging on the same underlying concern: how to curb entrenched platform power.

 

South Korea, India and Brazil: Additional international examples make the pattern even clearer. South Korea was the first country to curb app-store billing dominance through legislation aimed at preventing Apple and Google from forcing developers to use their in-app payment systems, and regulators later moved toward issuing fines for alleged non-compliance. In India, the Competition Commission has pursued Google over Play Store billing and related app-market practices, framing them as discriminatory and harmful to local developers and digital businesses. In Brazil, CADE has pushed Apple toward opening iOS to rival app stores and alternative payments, extending the global trend of using competition law to pry open tightly controlled mobile ecosystems. These examples matter because they show that concerns over gatekeeping, billing control, and technical lock-in are no longer confined to the EU or North America.

 

Australia and Canada: A second cluster of high-impact instruments focuses on platform–publisher bargaining and media-market power. Australia’s News Media and Digital Platforms Mandatory Bargaining Code created a regulated framework for commercial bargaining between designated digital platforms and eligible news businesses under ACCC oversight. Canada’s Online News Act framework is similarly being implemented through CRTC-supervised bargaining, mediation, and final-offer arbitration. These initiatives are not classic antitrust in form, but they clearly respond to the same structural concern: when a handful of platforms dominate traffic, advertising, and audience access, bargaining power in the information ecosystem can become severely distorted.

 

OECD: Finally, policy initiatives and institutions are also pushing for convergence to prevent enforcement from fragmenting across borders. The OECD and the International Competition Network (ICN) remain central venues for shared analysis on platform power, data, and algorithmic decision-making, helping competition authorities compare tools and align at least some core concepts even when their legal systems differ. What is emerging, in effect, is a global regulatory vocabulary built around ideas such as contestability, interoperability, data access, self-preferencing, and gatekeeper obligations. The laws differ, but the direction of travel is increasingly shared.

 

The role of intermediaries

In digital governance terms, competition regulation increasingly treats major platforms as intermediaries with rule-setting power. They can shape visibility (ranking and recommendation), access (app stores, operating systems, identity systems), and monetisation (adtech). 

That is why modern competition remedies often intersect with other policy domains: privacy (when data access is discussed), cybersecurity (when interoperability is mandated), consumer protection (when dark patterns affect switching), and even content governance (when recommender systems determine which publishers or voices can reach audiences). The overlap is increasingly the core reality of competition regulation in a digital and AI-mediated public sphere.