A. INTRODUCTION
The present era of digitalization, where artificial intelligence (AI) directly shapes the algorithmic business ecosystem, gives rise to a need for in-depth solutions regarding the impact of AI use in the field of competition law.
With the recent acceleration of digitalization, AI systems have become a powerful force determining behaviors not only in service and production sectors but across all market actors. Factors such as algorithmic pricing, data-driven forecasting and user guidance show that competition is shaped not only by humans but also by algorithmic models. These models diverge from classical competition law principles and trigger debates that redefine (i) intent, (ii) collusion, and (iii) abuse of dominance. Indeed, autonomous systems often make decisions beyond human supervision, thus becoming invisible coordinators of the economy.
This transformation shifts competition analysis from a “human behavior” perspective to a “machine behavior” perspective. Competition authorities must now analyze not only behavioral economics but also the algorithmic interaction dynamics emerging from digital transformation. The European Commission and the Organisation for Economic Co-operation and Development (OECD) note that AI systems’ unpredictable learning capabilities complicate traditional assessments of “intent to infringe” (OECD, Algorithms and Collusion: Competition Policy in the Digital Age, OECD Policy Roundtable Report, 2017).
B. OVERVIEW
AI has reshaped the actions considered to be competition law infringements. Algorithms now determine pricing decisions, supplier relations, and advertising rankings, giving rise to new forms of infringement such as “algorithmic collusion” and “tacit coordination”. In such cases, even without direct communication between undertakings, algorithms may act in parallel through similar learning models, effectively restricting competition. Institutions such as the European Commission, the UK Competition and Markets Authority (CMA) and the OECD have begun to recognize that algorithmic behavior can influence market outcomes regardless of intent. Thus, the question is no longer “Is there an infringement?” but rather “Who bears legal responsibility within the limits of algorithmic decision-making?”
The European Union’s Digital Markets Act (DMA) and Artificial Intelligence Act (AI Act) are among the first regulations to establish a systematic framework for this complex relationship. The DMA seeks to control algorithmic power by imposing obligations such as data-sharing, transparency, and a ban on self-preferencing on major platforms designated as “gatekeepers” (Regulation (EU) 2022/1925 of the European Parliament and of the Council (Digital Markets Act – DMA – https://eur-lex.europa.eu/eli/reg/2022/1925/oj/eng, accessed 01.10.2025).
The AI Act introduces ex ante oversight, explainability, and auditability requirements for high-risk AI systems (Proposal for a Regulation of the European Parliament and of the Council Laying Down Harmonised Rules on Artificial Intelligence (Artificial Intelligence Act), accessed 01.10.2025: https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=celex:52021PC0206 ).
Through these regulations, algorithmic decision-making processes are evolving into systems that are not only ethical but also accountable within the context of competition law. As a result, we are witnessing the emergence of a new liability regime that enables the law to limit data dominance in the market.
At the international level, the European Commission has issued numerous precedent-setting decisions addressing digital transformation and competition law. One of the most notable is the Google Shopping case (European Commission, Case AT.39740 – Google Search (Shopping), Decision of 27 June 2017.; accessed 05.10.2025: https://ec.europa.eu/competition/antitrust/cases/dec_docs/39740/39740_14996_3.pdf), which provides a concrete example of how algorithmic power can manifest as an abuse of dominance.
In this case, the European Commission fined Google €2.42 billion for engaging in self-preferencing, favoring its own comparison shopping service over competitors through ranking algorithms. This precedent established that algorithmic dominance in digital markets could no longer be viewed as a “natural competitive advantage” but rather as a “misuse of market power.”
In parallel with the international framework, the Turkish competition law is also adapting to new forms of infringement brought about by digitalization. Turkish Law No. 4054 on the Protection of Competition does not yet include specific provisions for algorithmic decision-making systems. However, existing clauses on “restriction of effective competition” and “abuse of dominance” may address the indirect consequences of algorithmic behavior. The Competition Authority’s 2021 report titled “Competition Policy in the Digital Economy” identified fundamental principles on algorithmic pricing, data portability, and platform behavior, raising awareness in this area. Still, for Türkiye to progress in alignment with the EU, it must strengthen institutional capacity in algorithmic transparency, data access and AI supervision and issue guidelines specifically addressing AI-based competition law infringements.
C. EUROPEAN COMMISSION DECISIONS AND ABUSE OF DOMINANCE
The concept of abuse of dominance refers to situations where an undertaking with substantial economic power acts independently of market conditions, unilaterally determining supply and demand dynamics and sustaining such conduct over time.
With advancing technologies, however, this concept has acquired new meanings in the digital economy. The monopolization of data access, manipulation of in-platform ranking algorithms and cross-use of user data now represent various and unprecedented incidents as digital counterparts of traditional “non-price competition” violations.
In a landmark investigation, the European Commission found that Meta’s (Facebook) data processing policies provided an unfair competitive advantage (European Commission, Case AT.40628 – Meta Platforms and Data Combination Practices, Ongoing investigation, 2023; accessed 05.10.2025: https://eur-lex.europa.eu/legal-content/EN/TXT/PDF/?uri=CELEX:62020TJ0451).
The Commission further emphasized that such conduct could constitute abuse under Article 102 of the Treaty on the Functioning of the European Union (TFEU), which prohibits dominant undertakings from exploiting their market power.
Additionally, the concept of “algorithmic coordination,” which can resemble concerted practices, is interpreted as effectively undermining competition when it results in outcomes such as parallel pricing without any collusion.
In the Eturas case (CJEU, Case C-74/14 – Eturas UAB and Others v. Lietuvos Respublikos konkurencijos taryba, Judgment of 21 January 2016: accessed 09.10.2025, https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=celex:62014CJ0074), the Court determined that an algorithmic price restriction on an online travel platform constituted indirect coordination. The decision demonstrated that algorithmic systems could function as “communication tools,” assigning liability not only to software developers but also to undertakings using such software.
The OECD’s 2017 report, “Algorithms and Collusion,” systematically analyzed the coordination risks inherent in algorithmic decision-making systems and recommended the adoption of ex ante oversight and algorithmic transparency principles.
D. ASSESSMENT UNDER TURKISH COMPETITION LAW
In parallel with the international framework, the Turkish competition law is also adapting to new forms of infringement brought about by digitalization. Turkish Law No. 4054 on the Protection of Competition does not yet include specific provisions for algorithmic decision-making systems. However, existing clauses on “restriction of effective competition” and “abuse of dominance” may address the indirect consequences of algorithmic behavior.
The Competition Authority’s 2021 report titled “Competition Policy in the Digital Economy” (https://www.rekabet.gov.tr/Dosya/dijital-piyasalar-calisma-metni.pdf, accessed 09.10.2025) identified fundamental principles on algorithmic pricing, data portability, and platform behavior, raising awareness in this area. Still, for Türkiye to progress in alignment with the EU, it must strengthen institutional capacity in algorithmic transparency, data access and AI supervision and issue guidelines specifically addressing AI-based competition law infringements.
As seen in the Trendyol (Competition Authority, Trendyol Decision No. 21-56/779-388, 30.09.2021; accessed 12.10.2025; https://www.rekabet.gov.tr/tr/Guncel/dsm-grup-danismanlik-iletisim-ve-satis-t-b3da5d0675adef1193d70050568585c9) and Yemeksepeti (Competition Authority, Yemeksepeti Decision No. 16-20/347-156, 09.06.2016; accessed 12.10.2025; https://www.rekabet.gov.tr/Karar?kararId=0bd0157a-2b4d-43ce-85a3-2af821bb387b) decisions of the Turkish Competition Authority, platform policies on algorithmic ranking and data use may be evaluated within the scope of the prohibition of “self-preferencing”.
These decisions indicate that Türkiye is undergoing a normative alignment process with EU regulations. Going forward, it is expected that structural transparency obligations, including the submission of algorithm audit reports to the Competition Authority, will be introduced through future legislative amendments to ensure effective oversight of algorithmic systems.
E. CONCLUSION
It is indisputable that digital transformation has introduced not only new forms of infringement but also new categories of liability, reshaping competition law. The definition of an “economic actor” is no longer limited to humans since autonomous decision-making systems have turned into actors that must be considered in legal analyses.
Accordingly, the future of competition law must be built on the principles of algorithmic transparency, accountability, and data sharing. In alignment with the European Commission and OECD guidelines, the preventive (ex ante) detection of algorithmic infringements is essential for protecting competition under Turkish law in the digital age.
Gülşah Güven, LL.M., Partner
