Introduction

Digital competition is set to remain a central priority for competition authorities worldwide. In Europe and beyond, authorities are increasingly shaping the business practices of large online platforms through the enforcement of competition law and dedicated digital competition regimes, particularly in the rapidly evolving artificial intelligence (AI) sector.

This forecast examines key market and regulatory developments expected in 2026. On the market side, AI is likely to continue transforming major digital markets, as firms both develop and deploy AI applications and increasingly acquire companies at early stages of development. On the regulatory side, competition authorities are expected to further adapt their approaches to digital markets, including by issuing new antitrust and merger control guidelines, designing or refining obligations for certain large online platforms, and responding more rapidly to emerging competition concerns.

Market Developments

AI markets are highly dynamic with frequent innovations at both the development (upstream) and deployment (downstream) stages of the value chain. While partnerships remain prevalent along the AI value chain to support these innovations, December 2025 has also seen a growing number of acquisitions.

Upstream Developments

At the upstream level, model developers continue to release new generations of models with enhanced capabilities. These include world models, which seek to capture physical and spatial aspects of the real world, and neurosymbolic models, which combine pattern recognition with rules and logic to support more structured decision-making. At the same time, developers increasingly rely on personal data provided by and generated from users, such as prompts, conversions, and outputs, to improve model performance at both the model and user levels[1]. This trend may give rise to data feedback loops, whereby increased user base and usage lead to improved performance, potentially conferring competitive advantages on widely used models[2].

Model performance is also improving rapidly, with frequent releases that outperform prior versions while reducing costs, as evidenced by competition among proprietary models such as OpenAI’s GPT series and Google’s Gemini models[3]. In parallel, open-source models, such as those developed by DeepSeek, are achieving performance levels comparable to proprietary alternatives, thereby intensifying competitive pressure on dimensions such as price, quality, and innovation.

These developments raise new research questions concerning the dynamics of competition in AI markets. In particular, the growing role of data, especially personal data, in model development, as well as the implications of data-driven feedback loops, warrants closer examination from both economic and policy perspectives.

Downstream Developments

At the downstream level, application developers continue to deploy and distribute AI services through various channels. On the consumer-facing side, model developers, such as OpenAI, are expanding their existing offerings with new features (e.g., shopping search, deep search, and agentic search), emerging ecosystems (e.g., application stores and e-commerce platforms), and complementary services (e.g., web browsers). Other developers, including Google and Meta, are pursuing alternative strategies by launching standalone AI services (e.g., Google Gemini) and integrating AI features into existing services, such as messaging services (e.g., Meta AI) or general search engines (e.g., Google AI Overviews).

On the business-facing side, proprietary model developers increasingly provide third-party access through application programming interfaces (APIs), enabling the development of a wide range of consumer and enterprise applications across sectors such as healthcare, legal services, and marketing. At the same time, intense competition and high operating costs of running models may prompt adjustments to business models, including changes to pricing, quality parameters, or monetisation strategies (such as the introduction of advertising)[4].

On both sides, developers increasingly focus their applications on AI agents that can perform tasks on users’ behalf, such as OpenAI’s Operator.

These developments call for renewed economic analysis of digital markets. AI-driven services are reshaping not only market boundaries and product characteristics, but also competitive interactions both within and across markets. In particular, the increasing use of AI agents has the potential to significantly alter the economics of the web, with implications for the economics of digital platforms and advertising as the web is shifting from human to agent-based interactions[5].

Acquisitions

Alongside partnerships, recent acquisitions in December 2025 across the AI value chain are beginning to shape market structure. At the upstream level, Nvidia entered into a $20 billion non-exclusive agreement with Groq, a firm developing custom chips for fast and cost-efficient inference that partially compete with Nvidia chips, while also acquiring key personnel, including Groq’s founder and president; Groq nevertheless remains operationally independent[6]. Google also acquired the energy provider Intersect for $4.5 billion, underscoring the strategic importance of securing new energy sources to power data centres for AI use[7].

At the downstream level, Meta acquired Manus for $2 billion—a company launched only in March 2025—to support the development of general-purpose AI agents across its consumer and business products, including Meta AI, whereas Meta is already investing significantly internally in AI[8].

These transactions underscore the need for careful analysis of how these acquisitions shape competition in AI-related markets, including energy, semiconductors, and AI agents. They are also likely to revive two major policy debates. The first concerns the assessment of pre-emptive acquisitions at an early stage of development, particularly where such transactions may escape review because they do not meet merger control thresholds. The second concerns the assessment of potential competitors under the “reverse killer acquisition” theory of harm, whereby an acquirer may discontinue or deprioritise its own innovation efforts following an acquisition, thereby reducing competitive intensity.

Regulatory Developments

Competition authorities continue to adapt their regulatory toolkits to address digital markets, combining traditional competition law enforcement with dedicated digital competition regimes.

Competition Law Enforcement

In the European Union, the Commission is expected to present new guidelines on abusive conduct by dominant firms (the revised Article 102 guidelines). Since the publication of the draft guidelines in August 2024[9], several important court judgments, such as Google Android Auto, have further shaped the interpretation of competition law in digital markets[10]. The final guidelines are therefore likely to reflect these developments, potentially relying more heavily on presumptions to define conditions under which certain practices may be considered abusive.

The Commission is also expected to publish draft guidelines on horizontal and vertical mergers, with particular relevance for digital mergers[11]. Over the past decade, merger assessments have increasingly accounted for factors such as multisided markets, non-price competition, data, and ecosystems, as illustrated by cases including Microsoft/LinkedIn, Apple/Shazam, and Booking/eTraveli.

Enforcement activity in AI-related markets is also accelerating. The Commission has opened antitrust investigations against Meta[12] and Google[13] in downstream markets and has signalled its willingness to intervene rapidly in the Meta case, including through interim measures. In parallel, Italy has already adopted interim measures in the same Meta case in December 2025, less than two months after Meta announced the alleged anticompetitive practice in October 2025[14].

Digital Competition Regimes

With respect to digital competition regimes, the EU will conduct its first review of the Digital Markets Act (DMA) by May 2026. The review will assess whether the DMA has achieved its objectives of contestability and fairness and whether it requires adaptation to new services or practices. Developments in AI are expected to play a significant role in this process, given that the Commission has explicitly sought feedback on the interaction between AI and the DMA[15].

In addition, the Commission is expected to publish a report on the possible introduction of new obligations in the cloud sector, following its first DMA market investigation into new practices[16]. This may prompt debate about the appropriate scope of the DMA, as the Commission is considering practices already covered by sectoral regulations (e.g., the Data Act) to enhance competitiveness rather than contestability[17].

Finally, in the United Kingdom, the Competition and Markets Authority is expected to propose conduct requirements for Google’s search and search advertising services, as well as for Google’s and Apple’s mobile ecosystems, and may initiate a market investigation into cloud services. These developments are likely to fuel discussions on international coherence among digital competition regimes.

This rapidly evolving market and regulatory environment underscores the need for continued research into the necessity, design, and proportionality of competition interventions to preserve the competition process without undue distortions in highly dynamic markets.

[1] Australian Competition and Consumer Commission, Recent Developments in Artificial Intelligence - Industry Snapshot, 17 December 2025 (accessed 5 January 2026). Available at: https://www.accc.gov.au/about-us/publications/recent-developments-in-ai-industry-snapshot

[2] Andrei Hagiu and Julian Wright, Artificial Intelligence and Competition Policy, International Journal of Industrial Organization, 2025 (accessed 6 January 2025). Available at: https://www.sciencedirect.com/science/article/pii/S0167718725000013

[3] Berber Jin, OpenAI Declares ‘Code Red’ as Google Threatens AI Lead, The Wall Street Journal, 2 December 2025 (accessed 6 January 2026). Available at: https://www.wsj.com/tech/ai/openais-altman-declares-code-red-to-improve-chatgpt-as-google-threatens-ai-lead-7faf5ea6

[4] OpenAI’s Cash Burn Will Be One of the Big Bubble Questions of 2026, The Economist, 30 December 2025 (accessed 7 January 2025). Available at: https://www.economist.com/leaders/2025/12/30/openais-cash-burn-will-be-one-of-the-big-bubble-questions-of-2026?utm

[5] Yingxuan Yang, Agentic Web: Weaving the Next Web with AI Agents, 2025 (accessed 7 January 2026). Available at: https://arxiv.org/abs/2507.21206

[6] David Faber, Nvidia Buying AI Chip Startup Groq’s Assets for About $20 Billion in its Largest Deal on Record, CNBC, 24 December 2025 (accessed 7 January 2026). Available at: https://www.cnbc.com/2025/12/24/nvidia-buying-ai-chip-startup-groq-for-about-20-billion-biggest-deal.html

[7] Alphabet Announces Agreement to Acquire Intersect to Advance U.S. Energy Innovation, Alphabet, 22 December 2025 (accessed 7 January 2026). Available at: https://abc.xyz/investor/news/news-details/2025/Alphabet-Announces-Agreement-to-Acquire-Intersect-to-Advance-U-S--Energy-Innovation-2025-DVIuVDM9wW/default.aspx

[8]  Angel Au-Yeung, Raffaele Huang, and Kate Clark, Meta Buys AI Startup Manus for More Than $2 Billion, The Wall Street Journal, 30 December 2025 (accessed 7 January 2026). Available at: https://www.wsj.com/tech/ai/meta-buys-ai-startup-manus-adding-millions-of-paying-users-f1dc7ef8

[9] Commission Seeks Feedback on Draft Antitrust Guidelines on Exclusionary Abuses, European commission, 1^st August 2025 (accessed 7 January 2026). Available at: https://ec.europa.eu/commission/presscorner/detail/en/ip_24_3623

[10] C-233/23 Alphabet e.a., ECLI:EU:C:2025:110, 25 February 2025.

[11] Review of the Merger Guidelines, European Commission (accessed 7 January 2026). Available at: https://competition-policy.ec.europa.eu/mergers/review-merger-guidelines_en

[12] Commission Opens Antitrust Investigation into Meta's New Policy Regarding AI Providers' Access to WhatsApp, European Commission, 4 December 2025 (accessed 7 January 2026). Available at: https://ec.europa.eu/commission/presscorner/detail/it/ip_25_2896

[13] Commission Opens Investigation into Possible Anticompetitive Conduct by Google in the Use of Online Content for AI Purposes, European Commission, 9 December 2025 (accessed 7 January 2026). Available at: https://ec.europa.eu/commission/presscorner/detail/en/ip_25_2964

[14] Autorità Garante della Concorrenza e del Mercato, A576 Meta AI: The Italian Competition Authority Orders Meta to Suspend the Terms Excluding Competing AI Chatbots from WhatsApp, 24 December 2025 (accessed 7 January 2026). Available at: https://en.agcm.it/en/media/press-releases/2025/12/A576

[15] Commission Gathers Views on How the DMA Can Support Fair and Contestable Digital Markets and AI Sector, European Commission, 27 August 2025 (accessed 7 January 2026). Available at: https://digital-markets-act.ec.europa.eu/commission-gathers-views-how-dma-can-support-fair-and-contestable-digital-markets-and-ai-sector-2025-08-27_en

[16] Commission Launches Market Investigations on Cloud Computing Services Under the Digital Markets Act, European Commission, 18 November 2025 (accessed 8 January 2026). Available at: https://ec.europa.eu/commission/presscorner/detail/en/ip_25_2717

[17] Christophe Carugati, The Politicisation of the Digital Markets Act, Digital Competition, 8 December 2025 (accessed 7 January 2026). Available at: https://www.digital-competition.com/comment/the-politicisation-of-the-digital-markets-act