The BRICS Competition Law and Policy Centre, in partnership with the HSE University Faculty of Computer Science's Scool of Data Analysis and Artificial Intelligence, convened a two-day seminar titled “AI and BRICS Competition Policy,” as part of the annual Artificial Intelligence and Society (AIS) Conference. The seminar brought together competition authorities, academic researchers, and technology companies from across the BRICS countries to examine two issues that are rapidly moving to the forefront of global competition policy: how artificial intelligence is reshaping market structures and competitive dynamics, and how AI itself can strengthen the enforcement capabilities of competition authorities. The second part of the Artificial Intelligence and Society Conference will take place from 29 June to 1 July.
Opening the seminar, HSE University Vice-Rector Alexey Koshel observed that policymakers around the world are increasingly asking how access to advanced AI capabilities may become restricted over time or be leveraged to facilitate unfair competition and monopolistic practices.
In the photo: Alexey Koshel © HSE University
He noted that, a year ago, China formally proposed establishing an international organization on artificial intelligence to prevent technological monopolization.
Chinese President Xi Jinping has articulated a key principle underlying this approach: artificial intelligence should be regarded as a global public good serving the interests of the international community. Echoing this broader debate, Russian President Vladimir Putin, speaking at the recent St. Petersburg International Economic Forum, argued that the previous model of global development has increasingly become an instrument of unfair competition. He called for a multipolar digital order and for fair competition among alternative technological solutions.
According to Maxim Shaskolskiy, Head of the Federal Antimonopoly Service (FAS Russia), artificial intelligence has become an integral component of the digital transformation of the economy. While AI delivers substantial benefits for businesses and consumers alike, it also creates a new generation of competition risks. Technologies designed to improve market efficiency can simultaneously reinforce the market power of dominant firms, while algorithms are increasingly capable of coordinating market behaviour more rapidly and effectively than humans.
Shaskolskiy outlined a number of initiatives being pursued by both FAS Russia and the international competition community. Among them, the authority is deploying digital tools to detect cartel conduct and expanding its state information system “Anti-Cartel”. As he noted:
"The regulator's task is not to fight technology, but to learn how to use it to protect competition. It may well be that we will need artificial intelligence itself to protect competition from the risks created by artificial intelligence."
In the photo: Maxim Shaskolskiy © HSE University
He also highlighted the ongoing work of the BRICS competition authorities to develop common approaches to AI governance and competition in digital markets, emphasizing the importance of international cooperation and striking an appropriate balance between safeguarding competition and preserving incentives for innovation.
"To improve our understanding of how artificial intelligence functions, assess existing regulatory approaches, and preserve fair competition in digital markets, the BRICS competition authorities are preparing a joint statement aimed at strengthening international cooperation. Together with the BRICS Competition Law and Policy Centre at HSE University, FAS Russia is also exploring deeper cooperation in the assessment of mergers in digital markets, including the development of a shared digital transaction-monitoring database — the so-called Merger Radar."
The Merger Radar initiative was first presented in Geneva in 2024 during the meeting of the BRICS Coordinating Committee on Antimonopoly Policy under Russia's BRICS Chairmanship and received support from competition authorities across the BRICS countries.
Alexey Ivanov, Director, BRICS Competition Law and Policy Centre, HSE Law School Professor, focused his remarks on how artificial intelligence is transforming market structures and reshaping the foundations of competition.
He argued that AI presents competition authorities with several structural challenges. Most fundamentally, AI technologies are altering the traditional balance between labour and capital, allowing a disproportionate share of the economic value they generate to accrue to a relatively small number of firms. At the same time, many leading AI companies continue to operate at a loss, a business model that may reflect long-term strategies aimed at securing market dominance and ultimately monopolizing emerging markets — an issue that is attracting growing attention from competition authorities.
In the photo: Alexey Ivanov © HSE University
Ivanov also pointed to an emerging tension between intellectual property regimes, which have traditionally enabled firms to capture returns on knowledge-based innovation, and a new AI-driven model of value creation that relies on the large-scale use of accumulated knowledge without necessarily recognizing or compensating underlying intellectual property rights. At the same time, he noted, AI markets are becoming increasingly concentrated, while a growing number of so-called strategic partnerships between major technology companies remain largely outside the scope of conventional merger review, despite their significant impact on market power and competitive dynamics.
According to Ivanov, the current wave of corporate concentration, combined with the widening technological gap between private firms and public regulators in the deployment of AI, requires competition authorities to pursue not incremental adaptation but a genuine technological and legal leap forward.
"The Merger Radar we are developing is essentially an AI agent designed to analyse merger activity and economic concentration on a global scale. It will enable competition authorities across our countries to respond more effectively to these emerging challenges. Without compromising confidentiality, such an AI system can transform fragmented market data collected by different competition agencies into actionable intelligence about competitive dynamics in global markets. Today, regulators largely see a static snapshot of markets. What we need instead is a moving picture. Effective competition enforcement requires a film rather than a photograph. If the BRICS countries succeed in building a shared legal and analytical framework for competition enforcement powered by advanced AI tools, it could significantly improve both the transparency and the competitiveness of the global AI marketplace."
Supporting the proposed Merger Radar initiative, Sarah Hassan Abdelhamid, Egyptian Competition Authority (ECA), emphasized that artificial intelligence is simultaneously making competition enforcement more challenging and more effective.
AI, she noted, introduces a range of novel enforcement challenges, including algorithmic coordination without explicit collusion, rapidly evolving digital markets, increasingly complex AI value chains, and growing concentration of critical resources among a small number of firms. At the same time, AI-powered analytical tools are fundamentally expanding the capabilities of competition authorities. Rather than relying primarily on complaints or investigating violations after they occur, regulators are increasingly able to adopt proactive market surveillance. Continuous monitoring systems can identify unusual market developments, detect anomalies, and flag potential competition concerns before they evolve into full-scale enforcement cases.
"At the ECA, we view AI as a transformative tool for competition enforcement. Through our Economic Informatics Unit, we have created data-driven systems that support market intelligence, price monitoring, bid rigging detection, and identification of potential competition concerns at an early stage. In many ways, AI is creating new blind spots for competition authorities, yet at the same time, it is giving us new tools to illuminate them. Our challenge is to ensure that our enforcement capabilities evolve at the same pace as the markets we are responsible for overseeing."
In the photo: Igor Drozdov © HSE University
Igor Drozdov, Deputy Chairman, VEB.RF, likewise argued that artificial intelligence presents a number of significant risks to competition, identifying five areas of particular concern.
First, he highlighted unequal and potentially restricted access to advanced AI systems, driven by both economic and geopolitical factors. Second, he pointed to data security and privacy, noting that the growing deployment of AI increases the potential for consumer manipulation and underscores the need to strike an appropriate balance between personalization and the protection of individual privacy.
Third, Drozdov warned of the growing market power of digital platforms, which, despite holding relatively few tangible assets, are able to achieve dominant positions through powerful network effects, creating heightened risks of market abuse. Fourth, he raised concerns about the displacement of original creative content by generative AI and stressed the importance of safeguarding the rights and economic interests of creators.
Finally, he addressed the role of mergers and acquisitions in the AI sector. While the concentration of capital and technological resources may be essential to support innovation in such a capital-intensive industry, it also carries the risk of excessive market concentration that could ultimately suppress competition.
"It is evident that AI is a highly capital-intensive sector, and some degree of resource concentration is both inevitable and necessary to drive innovation. The challenge for policymakers is to determine where that concentration ceases to promote innovation and instead begins to undermine competition by enabling firms to extract economically unjustified rents."
Fu Hongyu, Director, AI Governance Center, Data Economy Center, Alibaba Research Institute, examined the rapid evolution of AI agents, arguing that they represent the next stage of artificial intelligence beyond today's generative AI models. Unlike conventional generative systems, AI agents are capable not only of responding to user prompts but also of accessing external data sources, making decisions, and executing tasks autonomously and continuously, without human intervention.
According to Fu, this growing autonomy is giving rise to entirely new forms of market behaviour, including automated pricing, autonomous commercial decision-making, and increasingly sophisticated forms of algorithmic coordination.
In the photo (up on the screen): Fu Hongyu © HSE University
"Overall, we are moving towards what can be described as an agent economy. This raises a host of new questions for competition law. Should AI agents themselves be regarded as platforms? How should access to and use of data be regulated? What constitutes the unjustified appropriation of third-party data? How can regulators prevent autonomous algorithmic collusion? And ultimately, who should bear responsibility for the actions of AI agents?"
From a competition perspective, Fu identified several emerging risks. These include unequal access to advanced AI agents, heightened privacy concerns stemming from their extensive access to user data, and growing complexity in the application of intellectual property law, as AI agents rely on data not only for model training but also for performing real-world tasks.
"This raises a fundamental question of accountability: should responsibility rest with the developer, the technology provider, or the end user?"
Fu also devoted particular attention to the ethical implications of increasingly autonomous AI systems. As AI agents evolve into persistent digital companions, he argued, their influence extends well beyond productivity, shaping users' social interactions and emotional behaviour. One example currently attracting growing regulatory attention is the emergence of AI companions for children. Young users may develop strong emotional attachments to such systems, raising concerns about child safety, psychological well-being, and appropriate safeguards.
He noted that policymakers in countries including China and Brazil are already debating whether certain forms of AI companionship for children should be subject to regulatory limits. For example, while children may be prohibited from using smartphones at school, they are often permitted to wear smartwatches equipped with AI assistants—a technological development that is creating entirely new markets as well as new regulatory challenges.
Fu argued that technology companies must therefore act responsibly not only from a commercial perspective but also in fulfilling broader social responsibilities. At Alibaba, he explained, this approach is reflected in an internal governance framework known as Model Spec — a comprehensive policy document defining what AI agents are permitted and prohibited from doing. The latest version of the framework contains six overarching principles and 46 detailed operational rules, together with a formal mechanism for resolving conflicts between competing values. For example, where commercial objectives conflict with the protection of children, child safety takes precedence.
Jia Kai, Professor, School of International and Public Affairs, Shanghai Jiao Tong University, China, proposed viewing AI's impact on the economy and competition as a three-stage process of market transformation.
At the first stage, AI serves primarily as a complementary technology, enhancing today's platform economy — or what is often described as the attention economy. At the second stage, it fundamentally reshapes that model through the emergence of an agent economy, in which AI agents embedded in smartphones, super-apps, and other digital services become increasingly autonomous participants in economic activity. At the third stage, AI evolves beyond existing platforms to create entirely new ecosystems capable of restructuring markets and presenting novel challenges for competition authorities.
At the same time, Jia Kai argued that the increasing concentration of capital and technological capabilities is being accompanied by the rapid expansion of the open-source AI ecosystem, which could serve as an important counterweight to market concentration and become a key pillar of future competition policy.
"Although AI may appear to be an increasingly concentrated industry, open-source models are becoming ever more significant. Projects such as DeepSeek, MiniMax, and other open AI initiatives are evolving into global public goods. This creates a meaningful alternative to a highly concentrated market structure for artificial intelligence."
In the photo: Vasiliy Gromov © HSE University
Vasiliy Gromov, Professor, School of Data Analysis and Artificial Intelligence, HSE University, argued that artificial intelligence should not be equated solely with large language models or generative AI. He distinguished between three broad categories of AI systems: interpretable AI, whose decision-making logic can be understood by humans; non-interpretable AI, including large language models; and self-interpreting AI, capable of continuously revising its own internal representation of the world as it acquires new data. From a regulatory perspective, he noted, the first two categories are of particular importance.
Gromov also emphasized that AI development is taking place under conditions of persistent constraints on resources, time, and expertise. Technological progress is advancing so rapidly that specialist knowledge can become outdated within a matter of months. In this environment, he argued, competitive advantage will belong not to those investing exclusively in technology or human capital, but to organizations capable of integrating AI with expertise in the humanities, organizational design, and effective management.
He further stressed that AI should primarily be understood as an amplifier of existing organizational processes. While it has the potential to dramatically improve the efficiency of businesses and public institutions, it can equally magnify existing weaknesses — including bureaucracy, poor governance, and other systemic shortcomings. For that reason, successful AI adoption requires not only digital transformation and workforce training, but also resilient governance structures and contingency mechanisms to mitigate technological failures or excessive dependence on automated systems.
The seminar's second day also featured a presentation by Tang Liming, Deputy Head, Platform Economy Institute, Competition Policy and Assessment Center, State Administration for Market Regulation of China (SAMR). Cooperation between the BRICS Competition Law and Policy Centre and SAMR's research institute has been developing since 2024, when the two organizations signed a strategic partnership agreement.
© HSE University
The programme also brought together officials, advisers, and researchers from competition authorities across the BRICS countries, including the Competition Commission of India (CCI), the Competition Commission of South Africa, Brazil's Administrative Council for Economic Defense (CADE), and the Competition and Consumer Policies Branch of UNCTAD.
Participants agreed to continue the dialogue at the 23rd Session of the UNCTAD Intergovernmental Group of Experts on Competition Law and Policy, to be held in Geneva in July. The agenda will include a dedicated session on emerging challenges in digital markets, where Alexey Ivanov, Director of the BRICS Competition Law and Policy Centre, has been invited to present the seminar's principal findings and the policy proposals developed during the discussions.
