1. Introduction
Artificial intelligence ("AI") has rapidly emerged as a transformative force in today's digital economy, fundamentally reshaping market dynamics and competitive structures. This technological evolution has redefined competitive strategies across industries, prompting both opportunities for innovation and new regulatory concerns. At its core, AI refers to computer systems capable of performing tasks that traditionally required human intelligence1, such as data analysis, decision-making, and pattern recognition. Among its most significant applications are algorithms - structured sets of rules and computations designed to process vast amounts of data and generate optimized outcomes. These innovations have redefined competitive strategies, offering substantial gains in efficiency and consumer experience. However, they also raise complex regulatory concerns, particularly in the realm of competition law enforcement.
As algorithms have moved from theoretical constructs to operational cornerstones in digital markets, their influence over market outcomes has become both pervasive and increasingly complex. With growing access to pricing and behavioral data, these tools can now forecast future patterns, track competitor strategies, and dynamically adjust prices. While such advancements can enhance competition by enabling consumers to make better-informed choices; they also allow firms to monitor rivals and engage in conduct that suppresses competitive rivalry. Indeed, the growing accessibility of pricing data, especially in digital markets, has significantly enhanced algorithmic capabilities in forecasting demand and consumer behavior2; and while this increased transparency is theoretically beneficial to consumers, they also create fertile ground for tacit coordination.3 Firms can closely monitor competitors' price changes and swiftly adapt their strategies, potentially dampening price competition. In such data-rich environments, algorithmic tools may be deployed not only to optimize internal pricing strategies but also to respond to competitors in ways that reduce strategic uncertainty and deter deviation. This dual nature of algorithmic tools underscores the need for nuanced assessments under competition law to distinguish between pro-competitive efficiencies and anti-competitive practices.
Ongoing developments underscore the pressing need for competition authorities to carefully evaluate whether algorithmic conduct fosters efficiency or veers into collusive or exclusionary practices. A fundamental challenge lies in understanding how algorithms harness data to influence key competition parameters such as price, choice, quality, and innovation. While algorithms can be used to outperform competitors and enhance consumer welfare, they also have the potential to harm consumers through excessive pricing and coordination-facilitating behaviors. In markets characterized by business-to-consumer ("B2C") transactions, increasing price transparency enables algorithmic price tracking and adjustments, which may be exploited for anti-competitive purposes. Beyond pricing, competition concerns extend to broader issues including algorithmic ranking, self-preferencing, and AI-driven market monitoring factors that are reshaping competitive dynamics in unprecedented ways.
Competition concerns also arise in the context of mergers and acquisitions ("M&A") within the AI sector. Indeed, digital platforms and technology firms increasingly leverage AI tools to influence market dynamics, raising questions about exclusionary abuse and the deployment of AI for anti-competitive purposes. The Turkish Competition Authority ("TCA") and global regulators are intensifying their scrutiny of these developments, particularly in cases involving exclusionary practices, AI-sector mergers, and the use of AI tools in competition law enforcement. Recent decisions by the Turkish Competition Board ("TCB") and the European Commission ("Commission") underscore the growing focus on algorithmic practices and their impact on market competition, signaling a new era of regulatory vigilance.
2. Competition Law Concerns Arising from Algorithm-Driven Practices Under Article 4
In Turkish competition law practice, Article 4 of Law No. 4054 on the Protection of Competition ("Law No. 4054") plays a central role in addressing anti-competitive practice; defining agreements and concerted practices between undertakings, as well as decisions and practices of associations of undertakings, that prevent, distort, or restrict competition directly or indirectly in a particular market for goods or services as illegal. These anti-competitive agreements can occur either between competing undertakings (horizontal relationships) or between undertakings operating at different stages of production and distribution chain (vertical relationships). Regardless of their form or structure, such agreements are deemed unlawful if they have the potential to harm market competition, ultimately reducing consumer welfare and leading to inefficient resource allocation.
In this context, even coordinated actions or tacit collusion between undertakings, without a formal written agreement, may fall under the scope of Article 4. For a competition law violation to arise, the agreement must either restrict competition by its object or by its effect. Restrictions by object refer to practices that are inherently harmful to competition ("per se") and do not require a detailed assessment of their impact.4 A prominent example of such violations among competitors is cartels. On the other hand, restrictions by effect ("rule of reason") necessitate an evaluation of their actual impact on competition. Horizontal agreements, such as price-fixing, are typically considered per se violations, while vertical agreements, such as recommended retail prices, are usually assessed under the rule of reason. In the realm of algorithmic practices, some activities may constitute infringements by object, while others require an effects-based analysis to determine their competitive impact.
- Competitive Concerns of Algorithms in Horizontal Relationships
The rise of digitalization has transformed the competitive landscape by enabling more efficient price monitoring and coordination mechanisms, thereby significantly increasing the potential for cartels. In traditional markets, the TCB has identified cases where suppliers acted as a "hub," coordinating price transitions between competing retailers to create a "hub-and-spoke" cartel. These suppliers collected competitively sensitive information through retailers and intervened when undertakings reduced prices or refrained from increasing them, ensuring alignment across the market. In digital markets, however, such coordination becomes even more aggressive, as algorithms facilitate real-time monitoring and enforcement of price agreements, making deviations easier to detect and collusion harder to disrupt.
Algorithmic pricing in competitive relationships can be categorized into three distinct scenarios.5 In the first scenario, firms explicitly agree on prices and use algorithms to implement this agreement, which is unequivocally prohibited under competition law; while in the second scenario, firms rely on a common third-party algorithm to set their prices. This setup often resembles a hub-and-spoke cartel, where the algorithm provider acts as the hub, and the sellers (spokes) align their pricing strategies based on its guidance.6 The third and most complex scenario arises when firms independently use pricing algorithms that unintentionally align, resulting in supracompetitive prices without any explicit agreement or direct coordination. This third scenario raises particular concerns due to the autonomous coordination enabled by algorithms, which removes the need for direct communication between firms. Traditional cartel detection methods, reliant on evidence of explicit collusion, may prove ineffective here. Consequently, regulatory bodies must focus on economic impact analysis to assess whether algorithm-driven coordination has led to anti-competitive effects. The Göltaş decision7 exemplifies this approach, where sanctions were imposed based solely on economic analysis despite the absence of direct communication evidence. This case underscores the significant risks algorithms pose to market competition and highlights the necessity of vigilant regulatory oversight by the TCB.
This issue is particularly prevalent in oligopolistic markets, where a small number of interdependent firms dominate pricing and production strategies. Such interdependence allows firms to sustain high prices without direct competition, distinguishing oligopoly from other market structures like perfect competition or monopoly. With the growing use of algorithms, concerns are rising that coordinated behavior could extend beyond traditional oligopolies to a wider range of market structures. For instance, in the Trendyol decision8, the TCB examined whether the platform's algorithmic pricing tool posed a price-fixing risk. It found that Trendyol's automatic pricing mechanism aligned competing sellers' prices in real time, enabling tacit coordination without explicit agreements. The algorithm also functioned as a monitoring tool, instantly detecting price deviations and responding accordingly, thereby strengthening the sustainability of potential cartel arrangements. Similarly, in the Hepsiburada decision9, the TCB scrutinized the platform's automatic pricing mechanism and its potential to produce cartel-like outcomes. A key concern was the "Match Buybox Price" feature, which locked multiple sellers into the same price and discouraged individual price cuts, as any reduction would be immediately matched by competitors. Recognizing the difficulty of detecting algorithm-driven collusion in digital markets, the TCB accepted a set of commitments from Hepsiburada to mitigate these risks which included removing the price-matching function, ensuring sellers were not pressured to use the automatic pricing tool, and preventing the sharing of data about competitors' usage of the mechanism. These measures aimed to preserve the pro-competitive benefits of dynamic pricing while reducing the risk of anti-competitive practices.
The TCB's Trendyol and Hepsiburada decisions provide a detailed analysis of the anti-competitive risks posed by algorithms and outline measures to mitigate these concerns. Specifically, they highlight how algorithms can facilitate price-fixing, concerted practices, and the exchange of competitively sensitive information, thereby increasing the risk of collusion. By enhancing market transparency, algorithms allow firms to monitor competitors' pricing behaviours more efficiently, reducing strategic uncertainty and fostering tacit coordination. A comparable issue was addressed in the Commission's Container Shipping case10, where pre-announcements of future price increases were classified as concerted practices because they allowed shipping companies to align their pricing strategies. This case illustrates the challenges regulators face: while explicit communication was involved there, the rise of autonomous algorithms could enable similar coordination without direct contact, posing significant difficulties for existing legal frameworks to effectively detect and address such conduct.11
- Vertical Agreements and Competition Restriction
Vertical agreements - those between firms operating at different levels of the supply chain - have long been subject to scrutiny by the TCB under established case law.While certain vertical restraints can enhance efficiency, practices such as Resale Price Maintenance ("RPM"), exclusive agreements, and Most-Favoured-Nation ("MFN") clauses pose significant risks of restricting competition. The increasing integration of algorithms in pricing, monitoring, and enforcement has further complicated the TCB's assessment of these agreements, introducing new dimensions to how vertical restraints are implemented and controlled.
Among these practices, RPM remains one of the most contentious. Under Article 4 of Law No. 4054, RPM is treated as a per se violation due to its direct limitation on competition and adverse effects on consumer welfare. The Guidelines on Vertical Agreements emphasize that RPM restricts retailers' ability to compete on price, thereby undermining market dynamics. The advent of algorithms as sophisticated price-monitoring tools has intensified this concern by enabling suppliers to detect and enforce price restrictions with unprecedented precision.
Real-time monitoring facilitated by these technologies ensures stricter compliance, amplifying the potential harm caused by RPM. This growing reliance on advanced pricing tools is well illustrated by the Commission's Consumer Electronics decision12, which exemplifies how algorithms and similar technologies are increasingly used to enforce RPM and other vertical restraints. While these tools are not inherently anti-competitive, they significantly enhance firms' ability to detect and maintain price restrictions, raising important competition concerns.
As algorithms become more prevalent in enforcing vertical restraints, traditional detection methods employed by the TCB may no longer suffice. In cases involving RPM, MFN clauses, or exclusive distribution agreements, competition authorities must adopt more comprehensive and sophisticated approaches. Advanced economic analysis combined with robust monitoring mechanisms is essential for effectively assessing the role of algorithms in digital markets. By integrating these tools, regulators can more accurately identify and address anti-competitive practices, ensuring that vertical agreements do not undermine market competition or harm consumer welfare.
3. Competition Law Concerns Arising from Algorithm-Driven Practices Under Article 6
Article 6 of Law No. 4054 prohibits the abuse of a dominant position in the market, whether exercised independently, through agreements, or via concerted practices. A central focus under this provision is exclusionary abuse, which involves strategies aimed at driving competitors out of the market. Such conduct indirectly harms consumers by weakening competitive dynamics and limiting consumer choice. The increasing integration of algorithms into commercial processes has made exclusionary abuse more sophisticated and challenging to detect, presenting new obstacles for competition authorities. Among the most prominent manifestations of such abuse in digital markets are algorithmic ranking and self-preferencing. However, for these practices to constitute a violation under Article 6, the undertaking in question must hold a dominant position. Digital platforms such as Trendyol, Hepsiburada, Getir, and Sahibinden.com rely heavily on algorithms to rank products, content, and services, thereby directly influencing consumer choices. Given that users tend to trust platform rankings, these algorithmic decisions have become a critical factor shaping competition dynamics in digital markets.
In the Trendyol decision13, the TCB found that the platform had favored its own sellers over third-party sellers, leading to an administrative fine of approximately TRY 62 million. The investigation revealed that Trendyol's dual role as both a marketplace and a retailer enabled it to manipulate algorithms and use third-party data to prioritize its own products, thereby discriminating against other sellers. This self-preferencing behavior was deemed an abuse of dominance under Article 6 of Law No. 4054, highlighting the competitive harm caused by algorithm-driven practices in digital markets. Similarly, in the Google Shopping decision14, both the TCB and the Commission ruled that Google abused its dominant position in the general search market by prioritizing its own shopping service while demoting third-party comparison-shopping services. Specifically, the Commission highlighted that Google's Panda algorithm systematically reduced the visibility of competitors, creating an exclusionary effect on competition. This practice led to a significant increase in traffic to Google's own service and a corresponding decline for rival platforms, underscoring the powerful impact of algorithmic self-preferencing on market dynamics. Together, these cases exemplify the heightened scrutiny competition authorities are applying to algorithm-driven conduct that leverages market dominance to distort competition; demonstrating the need for vigilant regulatory oversight to ensure that digital platforms do not exploit their algorithmic capabilities to the detriment of competitors and consumers alike.
4. Competition Law Concerns Arising from Algorithm-Driven Practices Under Article 7
Article 7 of Law No. 4054 prohibits mergers, acquisitions, or similar transactions that result in a significant lessening of effective competition within a market, particularly through the creation or strengthening of a dominant position. This provision is particularly relevant to the AI-related industry, where rapid consolidation among technology firms raises concerns about monopolistic structures and reduced competition. A significant issue in this context is that many emerging technology firms traditionally fall below the standard turnover thresholds, which could create a regulatory gap.15 However, this challenge is addressed in the Turkish competition law practice through the amendment made in the Communiqué Concerning the Mergers and Acquisitions Calling for the Authorization of the Competition Board in 2022, which introduced a specific framework for technology undertakings, ensuring that acquisitions involving firms exceeding certain global turnover thresholds are subject to TCB review, regardless of whether these firms operate directly in Türkiye. This approach allows the TCB to effectively oversee transactions involving smaller but strategically important AI companies that might otherwise escape scrutiny under traditional turnover criteria.
Beyond traditional mergers and acquisitions, investments and strategic partnerships between major technology firms; such as the Microsoft-OpenAI16 collaboration, Google's investment in Anthropic, further complicate the competitive landscape. While these arrangements may not constitute conventional acquisitions subject to merger control, they have the potential to consolidate market power and reduce competition by enabling dominant firms to combine their resources, data, and expertise, which could raise significant entry barriers for smaller competitors.
Indeed, the Microsoft-OpenAI partnership, by integrating advanced AI technologies into Microsoft's cloud infrastructure and product ecosystem, has attracted regulatory scrutiny in jurisdictions including the United States17 and the United Kingdom18. Concerns focus on whether this collaboration may limit competitors' access to similar capabilities or create "lock-in" effects that hinder market entry. Similarly, Google's investment in Anthropic is viewed as a strategic move to strengthen its AI development capacity and consolidate valuable data and knowledge, potentially reinforcing its market position19. However, investigations and assessments by competition authorities are ongoing, and definitive conclusions have yet to be reached.20
5. TCA's AI-Supported Projects
In December 2024, the TCA announced21 its plans to deploy AI-based tools to detect price anomalies, particularly in the food and service sectors. Under this initiative, leading indicators will be identified using data obtained from the Turkish Statistical Institute, and machine learning algorithms will be employed to predict potential price anomalies in advance. This proactive approach aims to enhance the TCA's ability to monitor markets and intervene swiftly when irregularities are detected.
As part of its AI-driven efforts, the TCA is also strengthening its collaboration with the Public Procurement Authority to develop analytical tools powered by AI. These tools are designed to detect anti-competitive risks and bid-rigging in public tenders. By leveraging advanced data analysis, the initiative seeks to improve transparency in procurement processes, proactively identify potential violations, and mitigate anti-competitive practices.22 This collaboration underscores the TCA's commitment to integrating cutting-edge technologies into its enforcement mechanisms.
Key consumer sectors, such as red meat, poultry, eggs, and retail, will be continuously monitored using AI-powered tools. This real-time monitoring enables rapid intervention when necessary, ensuring that market conditions remain fair and competitive. These initiatives are part of the TCA's broader digitalization strategy, which aims to enhance competition law enforcement by leveraging AI and machine learning technologies across various markets. From consumer goods to public procurement, these tools are expected to play a pivotal role in maintaining competitive market dynamics.
6. Conclusion
The increasing integration of AI and algorithms into markets has brought both significant efficiency gains and new challenges for competition law enforcement. While these technologies optimize market operations and enhance decision-making, they also pose risks related to price-fixing, tacit collusion, and exclusionary practices. Specifically, the ability of algorithms to process vast amounts of data and react instantaneously raises concerns about their potential impact on market competition, particularly in pricing and monitoring mechanisms. Therefore, it is crucial to assess whether these technologies facilitate behaviors that harm competition, such as price agreements or exclusionary conduct.
In horizontal relationships, for example, algorithmic coordination can stabilize prices and reduce uncertainty among competitors, potentially leading to anti-competitive outcomes even in the absence of explicit agreements. Similarly, in vertical relationships, the use of AI-driven tools for pricing enforcement and market monitoring has raised concerns about practices like RPM. These developments highlight the need for a comprehensive assessment of the role of algorithms in competition law enforcement. Moreover,beyond anti-competitive agreements, algorithms are also relevant in abuse of dominance cases, particularly regarding self-preferencing and algorithmic ranking. Digital platforms that rely on AI-driven mechanisms for product visibility and ranking have faced scrutiny over whether such practices restrict competition. Additionally, mergers and acquisitions in the AI sector presents specific challenges, as certain transactions involving technology firms may not meet traditional notification thresholds. This requires competition authorities to adopt more nuanced approaches to assess their impact.
In response to these evolving challenges, competition authorities are progressively enhancing their enforcement strategies by supplementing traditional detection and assessment methods with advanced economic analyses and digital monitoring tools. Within this context, under Turkish competition law practice, the TCA's AI-based initiatives exemplify a broader effort to strengthen enforcement capabilities through data-driven approaches, particularly in areas such as price monitoring and bid-rigging detection. As AI's role in markets continues to expand, it is imperative that competition law enforcement evolves accordingly to safeguard competitive conditions and protect consumer welfare.
Footnotes
1 See, https://builtin.com/artificial-intelligence#:~:text=Artificial%20intelligence%20refers%20to%20computer,speech%20and%20generating%20natural%20language.
2 OECD, Algorithms and Collusion: Competition Policy in the Digital Age, 2017, see, https://www.oecd.org/daf/competition/Algorithms-and-colllusion-competition-policy-in-the-digital-age.pdf,
3 EZRACHI / STRUCKE, Artificial Intelligence & Collusion, s.1781.
4 Decision dated 07.03.2011 and numbered 11-13/243-48 of TCB.
5 These three scenarios are in line with Autorité de la concurrence and Bundeskartellamt, "Algorithms and competition" (Report) 2019, 26–60 and Autoridade da Concorrência, "Digital ecosystems, big data and algorithms" (Issues Paper) 2019,
6 Advocate General Spuznar already suggested the hub-and-spoke qualification for Uber in Case
C-434/15 Asociación Profesional Elite Taxi v Uber Systems Spain EU:C:2017:364,
. Another potential qualification is that of cartel facilitator, as in Case C-194/14 P AC-Treuhand v Commission EU:C:2015:717, "but that qualification appears more suited to firms (such as consultan- cies) that operate on a completely different market".
7 Decision dated 20.09.2007 and numbered 07-76/908-346 of TCB.
8 Decision dated 03.10.2024 and numbered 24-40/950-409 of TCB.
9 Decision dated 03.10.2024 and numbered 24-40/951-410 of TCB.
10 Container Shipping (Case AT.39850) Commission Decision of 7 July 2016. Note that the case ended
with commitments so there is no final decision, let alone a judgment confirming it.
11 The Cambridge Handbook of The Law, Ethics And Policy Of Artificial Intelligence, section 9, Artificial Intelligence and Competition Law.
12 European Commission - Press Release Database: "Antitrust: Commission fines four consumer electronics manufacturers for fixing online resale prices", Jul 24, 2018, https://ec.europa.eu/commission/presscorner/detail/en/ip_18_4601
13 Decision dated 23.09.2021 and numbered 21-44/650-M of TCB.
14 Decision dated 13.02.2020 and numbered 20-10/119-69 of TCB.
15 OECD, Artificial Intelligence, Data And Competition, 2024, see, https://www.oecd.org/en/publications/artificial-intelligence-data-and-competition_e7e88884-en.html
16 See, https://www.reuters.com/world/uk/uk-antitrust-regulator-considering-microsoft-openai-partnership-2023-12-08/
17 https://www.forbes.com/sites/jackkelly/2024/06/07/us-regulators-are-looking-into-microsoft-nvidia-and-openai/; https://www.reuters.com/technology/microsoft-ditches-openai-board-observer-seat-amid-regulatory-scrutiny-2024-07-10/
18 https://www.gov.uk/cma-cases/microsoft-slash-openai-partnership-merger-inquiry
19 https://www.nytimes.com/2025/03/11/technology/google-investment-anthropic.html
20 Federal Trade Commission, "FTC Launches Inquiry into Generative AI Investments and Partnerships", 25 January 2024, https://www.ftc.gov/news-events/news/press-releases/2024/01/ftc-launches-inquiry-generative-ai-investmentspartnerships
21 See, https://haber.rekabet.gov.tr/haber/rekabet-sorusturmalarina-yapay-zeka-destegi
22 See, https://www.rekabet.gov.tr/tr/Guncel/kamu-ihale-kurumu-ve-rekabet-kurumu-kamu-1c9a93d8b5e9bef1193d20050568585c9
The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.
