CYIL vol. 11 (2020)

CYIL 11 (2020) ARTIFICIAL INTELLIGENCE AND COMPETITION LAW… as far as pricing is concerned; this topic was outlined by professors Ezrachi and Stucke, 7 discussed on prominent international fora, including the Organization for Economic Cooperation and Development (OECD) 8 and lately addressed by reports of leading competition authorities, including the UK’s report at the end of 2018 (UK Study) 9 and most recently a joint report of French and German competition authorities in November 2019 (Franco-German Study). 10 Even though the importance of other algorithms for competition policy cannot be excluded, we will only focus further on pricing algorithms , i.e. algorithms “ that uses price as an input, and/or uses a computational procedure to determine price as an output ”; 11 these include price monitoring algorithms, price recommendation algorithms, and price-setting algorithms. 12 For the purposes of this article, we will discuss price-setting algorithms using dynamic pricing . i.e. changing the prices on the basis of the costs, capacity, or demand; 13 especially in connection with the abuse of dominance, we will also discuss personalised pricing algorithms , i.e. algorithms charging different prices to consumers based on their personal characteristics. 14 1. Dynamic Pricing Algorithms Dynamic pricing algorithms autonomously set the prices, while taking into account different variables. The most prominent example of this strategy is the airline industry, which has been employing it for decades. 15 From the point of view of competition policy, it is important to realise that dynamic pricing algorithms may work not only with “internal” data of the undertaking in question (production costs, stocks, demand etc.), but also prices set by their competitors. According to the study performed by the European Commission, one half of online sellers systematically monitor prices set by their competitors, more than two thirds of them using specific applications; more than three quarters of them then use this information to adjust their own We will in particular refer to the followingOECD’s publications: ORGANISATIONFOR ECONOMICPOLICY AND DEVELOPMENT. Algorithms and Collusion: Competition Policy in the Digital Age (2017), available at: http://www.oecd.org/daf/competition/Algorithms-and-colllusion-competition-policy-in-the-digital-age.pdf (1 July 2020), OECD Report on Algorithms, and ORGANISATION FOR ECONOMIC POLICY AND DEVELOPMENT. Personalised Pricing in the Digital Era. Background Note by the Secretariat (2018), available at: https://one.oecd.org/document/DAF/COMP(2018)13/en/pdf (1 July 2020), OECD Report on Personalised Pricing. 9 COMPETITION & MARKETS AUTHORITY. Pricing Algorithms. Economic Working Paper on the Use of Algorithms to Facilitate Collusion and Personalised Pricing (8. 10. 2018). Available at: https://assets.publishing. service.gov.uk/government/uploads/system/uploads/attachment_data/file/746353/Algorithms_econ_report. pdf (1 July 2020). 10 AUTORITÉ DE LA CONCURRENCE AND BUNDESKARTELLAMT. Algorithms and Competition (2019), available at: https://www.bundeskartellamt.de/SharedDocs/Publikation/EN/Berichte/Algorithms_ and_Competition_Working-Paper.pdf?__blob=publicationFile&v=5 (1 July 2020). 11 UK Study , par. 2.4. 12 Ibid , par. 2.5. 13 According to OECD Report on Personalised Pricing, par. 18, dynamic pricing means “ adjusting prices to changes in demand and supply, often in real time, not implying any discrimination between consumers” . 14 Ibid . 15 For example, the American Airlines introduced such an algorithm already in 1968. In detail, see Franco-German Study , p. 5. 7 EZRACHI, A., STUCKE, M. E. Virtual Competition. The Promise and Perils of the Algorithm-driven Economy. Harvard University Press, 2016. 8

241