Using AI makes business more effective. However, it also increases the number of actual and potential disputes about violations of confidentiality and unfair competition, as well as breaches of laws on advertising. Currently, existing legal provisions are insufficient to ensure that state-of-the-art technologies are safely implemented and used. Let us consider what antimonopoly risks are associated with the use of AI and what legal instruments exist for protection in Russia and abroad.
Essentially, AI means that a machine is able to perform tasks that usually require human intellect. The power of these technologies is fascinating: in a matter of minutes, such a computer program analyses gigantic volumes of data. This, in turn, helps companies enhance their business processes and improve performance. In addition, businesses that use AI can create personalised products and services, improving customer satisfaction and increasing their competitive power.
For example, banks and financial institutions use AI to analyse credit scores and online retailers use machine learning algorithms to analyse the past purchases of clients to offer them the most suitable products.
The main technological trend is the development and broad expansion of generative AI, a powerful instrument for creating new content, including images, videos and so on. Generative models can speed up the process of producing content and solving creative tasks. They can also learn successfully based on vast amounts of information.
Using AI gives rise to questions connected with copyright, data confidentiality and parties that are subject to liability
However, using AI gives rise to questions connected with copyright to the content created with AI, data confidentiality, what parties should be subject to liability and so on. In view of this, additional regulation is necessary of how digital technologies are implemented and used.
Several approaches have been formed to regulating relationships in connection with AI.
European Union At the EU level, certain legal provisions are expected to be adopted that will regulate the development, implementation and use of AI systems. For example, the Artificial Intelligence Act1 was adopted in 2024 which aims at ensuring security when AI is used.
China Together with some other countries, China keeps to a hybrid approach combining elements of hard and soft regulation, as well as self-regulation. Please note that this country is very active in implementing legal strategies and mechanisms. In 2023, the world’s first Interim Measures for the Administration of Generative Artificial Intelligence Services2 came into force. The rules set out in this document prohibit the generation of content connected with subverting the state's power, promoting terrorism, and so on.
Singapore The approach used in this country includes regulation based on soft law instruments only. The reference points for the policy connected with AI-related law and ethics are set by the Model Artificial Intelligence Governance Framework3 which, among other things, provides that models need to be checked regularly to ensure that they work correctly and to eliminate systemic mistakes.
Russia In Russia, a comprehensive system is developing of regulating public relationships connected with AI. It includes legal, technical and ethical regulation.
Additional regulation is needed of how AI is implemented and used
The first and most important legal instrument in this field was the Russian President’s Decree No. 490 dated 10 October 2019 “On developing artificial intelligence in the Russian Federation”. This decree approved the National Strategy for the Development of AI until 2030.
Let us look at what risks of breaching antimonopoly and advertising legislation the use of AI can create.
Deepfakes
Deepfakes are videos in which, using machine learning algorithms, one person’s face is replaced with another's. Lately, such deepfakes have become increasingly popular in advertising.
Many of you could see a commercial of Megafon starring the Hollywood actor Bruce Willis4, or, more precisely, his digital copy. You could also see a commercial of Tula honey-cake starring (a fake) Keanu Reeves5.
Disseminating deepfakes increases the risks of information being manipulated and inaccurate information being disseminated.
According to Federal Law No. 38-FZ “On advertising” dated 13 March 2006, advertising must be good-faith and accurate. At the same time, there is no restriction at the statutory level on the use of deepfake technologies in advertising. However, if such advertising results in the consumer being mislead and provided with incorrect information, one can confidently say that the law has been breached.
Please also note that using deepfake technologies to perform an action (which does not aim at advertising a product or service) which misleads consumers can be classified as a violation in the form of bad-faith competition under article 14.2 of Federal Law No. 135-FZ “On protecting competition” dated 26 July 2006.
In Russia, a comprehensive system is developing for regulating public relationships connected with AI
Pricing algorithms
A pricing algorithm is software which automatically sets prices in response to different changes, for example, in competitors’ prices, customer preferences and so on. Obviously, the bulk of such algorithms are currently based on AI.
Using pricing algorithms is not prohibited in itself unless it entails violating antimonopoly legislation, such as concluding “vertical" agreements, agreeing actions, coordinating business activities and setting up cartels.
Please remember that the Russian Federal Antimonopoly Service has already initiated cases against Samsung and LG when they coordinated resellers’ activities using pricing algorithms (the judgments in cases No. 1-11-18/00-22-17 dated 26 February 2018 and No. 1-11-11/00-22-19 dated 7 May 2019).
Personalisation models
Many services use personalisation models: a forecast is made based on the user’s history of contacting the service of what songs, movies or products they will prefer or will suit them best.
For example, Kinopoisk uses AI to recommend movies and serials to users.
In this situation, we can talk about targeted advertising aimed at those who are potentially interested in a product or service. In view of this, when using AI, it is necessary to take into account the requirements of legislation on advertising.
We believe that, with the development of AI technologies, growth should be expected in the number of legal disputes connected with different aspects of how they are used. At the same time, only a combination of different instruments and flexible approaches both at national and international levels can ensure proper regulation of how artificial intelligence is implemented and used.
1. Artificial Intelligence Act: MEPs adopt landmark law. URL: https://www.europarl.europa.eu.
2. Interim Measures for the Administration of Generative Artificial Intelligence Services. URL: https://www.cac.gov.cn.
3. Model Artificial Intelligence Governance Framework. URL: https://ai.bsa.org.
4. Bruce Willis has become a part a Russian mobile provider's advertising blockbuster // RBC. 16 August 2021. URL: https://style.rbc.ru.
5. Video: Keanu Reeves in an advertisement for Tula honey-cake // RG.ru. 12 September 2022. URL: https://rg.ru.
Главная цифра
315
дел об антиконкурентных соглашениях и координации экономической деятельности ФАС России возбудила в 2024 г.