WIXLIB

Uberlândia. Colunista do portal Magis. Atua como coordenadora da área deDireito Digital do Mansur Murad Advogados. E-mail: pietraquinelato@gmail.com.Rubens Cantanhede Mota Neto - Graduado em Direito pela Universidade deBrasília, membro do Grupo de Estudos Constituição, Empresa e Mercadocoordenado pela professora Ana Frazão.Sergio L. B. F. Reis - Assessor da Conselheira Lenisa Prado no CADE.Especialista em Regulação. Mestre em Direito pela Universidade CândidoMendes, graduado em direito e pós-graduado em propriedade industrial pelaUERJ.11

SUMÁRIOApresentação . 14A new language for A.I. and the legal discourse (Eduardo Molan Gaban Vinícius Klein). 16Private enforcement of privacy breaches from Big Tech Companies under theGDPR, Competition law provisions and the Unfair Commercial PracticesDirective: substitutable or complementary legal grounds? (Jeanne Mouton) . 44As plataformas digitais e os novos contornos do problema dos conglomerados:Novos critérios para a análise de atos de concentração e a necessária mudançade paradigma do direito concorrencial (André Luís Alvarenga Portella) . 72Competition and Regulatory Challenges in Digital Markets: How to tackle theissue of self-preferencing? (Frédéric Marty) . 109Aspectos concorrenciais dos preços personalizados em plataformas digitais –Uma análise comparativa (Pietra Daneluzzi Quinelato) . 149Compliance concorrencial no mercado digital (Ana Cristina Gomes Ana ElisaBertolin da Silva Gabriel de Aguiar Tajra) . 168Compartilhamento de dados pessoais intragrupo econômico: a (in)suficiência doconsentimento como base legal e desafios para o Direito Antitruste (MônicaTiemy Fujimoto) . 187Instituições regulatórias e desenvolvimento no ambiente digital (Gabriel deAguiar Tajra) . 212Streaming, consumidor e concorrência (Sergio L. B. F. Reis Lenisa RodriguesPrado Luiz Guilherme Branco) . 235Rastreabilidade da produção agrícola através da tecnologia: Impactosambientais e econômicos (Larissa Milkiewicz Aron Vitor Fraiz Costa). 26412

Machine learning: Regulatory control of stock manipulation in social media(Mariana Urban) . 289Impactos concorrenciais do Pix no sistema financeiro (Ana Paula Guimarãesde Paula Luísa Campos Faria) .312Open Banking in the light of law & economics: Reduction of informationalasymmetries to foster competition (Eduardo Oliveira Agustinho DiogoKastrup Richter) . 331A análise concorrencial e a COVID-19: Os efeitos da pandemia na atuação doCADE (Giovana Vieira Porto Julia Gonçalves Braga) . 357O direito concorrencial em tempos de crise: Crisis failing e firm defense (MatheusDalta Pimentel) . 386A crise do Consumer Welfare Standard (Rubens Cantanhede Mota Neto). 409Trocas de informações sensíveis sobre termos e condições de trabalho no direitoantitruste brasileiro (Fernanda Lopes Martins) . 441Direito Antitruste e Big Data: a conexão da proteção e dados com o direito daconcorrência (Alaís Aparecida Bonelli da Silva José Matheus Muniz Mariana Nascimento Silveira). 471O direito à portabilidade de dados: Obrigações e limites impostos a concorrentes(Isabela Maria Rosal Santos Luísa Campos Faria) . 501Aspectos concorrenciais na implementação do 5G no Brasil (Isabela MariaRosal Santos Luísa Campos Faria) . 517Regulation in the digital economy. Is ex-ante regulation of “gatekeepers” anefficient and fair solution? (Christophe Carugati) . 53513

ApresentaçãoA presente obra é resultado do 3º Congresso Internacional do Instituto Brasileirode Concorrência e Inovação – IBCI e traz os melhores artigos selecionados nachamada de artigos do evento, bem como trabalhos dos palestrantes do eventorelacionados aos seus painéis e pesquisas apresentadas.O leitor contará reflexões de grandes professores brasileiros e estrangeiros, bemcomo de promissores pesquisadores de várias origens e linhas de pesquisas.Este livro, em linha com o 3º Congresso Internacional do IBCI, atribuiu foco aosprinciaís desafios postos às autoridades públicas de todo o munfo pelo fenômenodisruptivo dos mercados e ecossistemas digitais.A heterogeneidade geográfica e de olhares para o avanço dos mercados digitaismesmo que do ponto de vista concorrencial demonstra como aspectosinterdisciplinares e interinstitucionais são relevantes para a compreensão eabordagem do tema.Os mercados digitais trouxeram o debate concorrencial de volta ao espaçopúblico e os desafios inerentes a essa realidade são complexos e podem resultarem mudanças estruturais não apenas em marcos legislativos, guias e diretrizesde defesa da concorrência, mas também na forma pela qual o processoconcorrencial acontece na sociedade atual.O enfrentamento desse contexto de falta de encaixe entre os fundamentostradicionais da defesa da concorrência e a nova realidade representa umaoportunidade de aperfeiçoamento da atuação estatal não apenas para osmercados digitais, mas de forma ampla. Por outro lado, uma resposta excessivado antitruste ou da intervenção estatal regulatória como resultado de umestranhamento com os novos processos de mercado e modelos de negócio podeimplicar risco à inovação não apenas no momento presente, mas tambémeliminar ciclos futuros de oportunidades e avanços.A relevância dessas questões tem feito com que diversas entidades tenham sedebruçado sobre o assunto e o IBCI apresenta esta obra como contribuição paraeste grande debate de forma interdisciplinar e plural e partir de uma abordagemacadêmica de excelência, valores estes que também guiam o IBCI.Eduardo M. Gaban / Vinícius Klein14

PrefaceThis book is the result of the 3rd IBCI International Conference on Competitionand Innovation and brings the best articles selected in the event's call for articles,as well as papers by the event's speakers related to their panels and researchprojects tackled in the Conference.The reader will access reflections from great Brazilian and foreign Professors,Scholars, as well as promising researchers from various origins and lines ofresearch. This book, in line with the 3rd IBCI International Conference, focusedon the main challenges posed to public authorities around the world by thedisruptive phenomenon of digital markets and digital ecosystems.The geographic heterogeneity and views of the advancement of digital markets,even from a competitive point of view, demonstrates how interdisciplinary andinterinstitutional aspects are crucial to the understanding and approach of thematter.Digital markets have brought the competitive debate back to the public arena andthe challenges inherent in this reality are complex and can result in structuralchanges not only in legislative frameworks, guides, and guidelines for the defenseof competition, but also in the way in which the competitive process happenstoday.Confronting this context of lack of fit between the foundations of competitionpolicy and the new reality represents an opportunity to improve the state'sperformance not only for digital markets, but in a broader way. On the other hand,an excessive antitrust response or regulatory state intervention in response to anestrangement from the new market processes and business models can imply arisk to innovation not only in the present moment, but also eliminate future cyclesof opportunities and advances.The relevance of these issues has meant that several entities have addressedthe subject and IBCI presents this work as a contribution to this great debate inan interdisciplinary and plural way, based on an academic approach ofexcellence, values that also guide IBCI.Eduardo M. Gaban / Vinícius Klein15

A new language for A.I. and the legal discourse1Eduardo Molan GabanVinícius KleinAbstract: The digital economy presents several challenges to antitrust lawenforcement. The traditional legal approach to define liability, as well as thetraditional enforcement toolkit are no longer effective. This context is present inthe case of algorithmic collusion. The central role of the concept of agreementand the focus on the results can still be effective to detect collusions outsidedigital markets, as digitalization is increasingly present in markets. On the otherhand, artificial intelligence (A.I.) and blockchains make virtually impossible theidentification of the authorship of crimes or offenses and pose difficulties for thetraditional normative matrix. This paper aims to address this issue of algorithmiccollusion in an innovative way and the use of A.I in favor of antitrust enforcement.The proposed solution to recover law effectiveness would comprehend acombination of antitrust and compliance by design in programming, producing amutation in the DNA Competition Policy, expanding, and modernizing the antitrustlaw enforcement toolkit and assuring its effectiveness in the digital markets.Keywords: Digital economy. Artificial intelligence. CompetitionCompliance by design. Liability. Effects. Authorship. Presumptions.policy.IntroductionAfter decades of quietude, antitrust has become “sexy again” 2. Thedebates are more explicit and profound, and they are not restricted to theclose community of experts but include the public opinion forum. Theemerge of digital markets is the main ingredient of this second revival ofthe antitrust movement3 and now the debate is not only restricted todifferent perspectives of Law and Economics but also technology issuesThis paper was selected by the Academic Society for Competition Law (ASCOLA) as one of the best papersin 2021 related to antitrust and digital economy, and it was presented by the authors in the Panel I –Algorithms and Antitrust, in the 16th annual event of Ascola.2SHAPIRO, Carl. Antitrust in a time of populism. International Journal of Industrial Organization, 61 (C),2018, p. 714.3According to HOFSTADER the Antitrust Movement in USA reached his highest point in 1914 with theClayton Act and the creation of the Federal Trade Commission, but after this progressive era and until1937 there as a phase of neglect, followed by a first revival in 1938 with the New Deal, but this time withno strong public sentiment against big business as in the progressive era. In the 1960s the antitrustmovement had faded away again. (HOFSTADTER, Richard. What Happened to the Antitrust Movement?In: HOFSTADTER, Richard. The Paranoid Style in American Politics and other essays. New York: VintageBooks, 2008, p. 193-204).116

and debates. Due to an intense transformation in the business environmentand in the competition process itself this rich environment of newchallenges has been accompanied by new perspectives and theories, suchas the New Brandeis Movement or Hipster Antitrust4. The most salientchallenge produced by this new business environment is the process ofcompetition5. In this specific topic, a possible adjustment of the antitruststandard of analysis and the toolkit used in the enforcement will benecessary to tackle the new reality of digital markets. Several reports aboutthe challenges of the digital market economy have been issued by antitrustauthorities and international actors.The presence of algorithms6 in our daily lives has severalrepercussions for antitrust analysis, but it is the use of machine learningthat has made the question of collusion more problematic as will beexplored in this paper. The discussion of the topic will make use of theframework presented in STUCKE & ERZACHI7, with four possiblescenarios of algorithmic pricing in collusion: messenger, hub and spoke,predictable agent and autonomous machine.In the first two scenarios the algorithm may be just a method by whichcollusive behavior is implemented. In the messenger scenario there is anagreement made by humans and the algorithm is just the messenger,executing orders such as monitor prices or punish deviants8. In the “huband spoke” collusion algorithms are a central hub to coordinate4These movement, despite its heterogeneity, has it its core the dissatisfaction with the consumer welfarestandard developed by the Chicago School. A starting point is: Khan, Lina. Amazon Antitrust Paradox.Yale Law Journal, 126 (3), pp. 710-805, 2016.5According to the UK CMA Working Paper on Pricing Algorithms, “An algorithm is any well-definedcomputational procedure that takes some value, or set of values, as input and produces some value, or setof values as output.”6An algorithm can be defined as: “a computer algorithm is a set of steps to accomplish a task that isdescribed precisely enough that a computer can run it.” Cormen, Thomas H. Algorithms Unlocked.Cambridge: MIT Press, 2013, p. 2.7EZRACHI, Ariel; STUCKE, Maurice. Artificial Intelligence & Collusion: When Computers InhibitCompetition. In: University of Illinois Law Review, 2017, pp. 1775-1810.8“UNDER OUR FIRST collusion scenario, Messenger, humans are the masters who agree to collude andmap out the cartel. The computer algorithms are the messenger, which the cartel members program to helpeffectuate the cartel and monitor and punish any deviation from the cartel agreement. (Ezrachi, Ariel andStucke, Maurice E. Virtual Competition: the promise and perils of the algorithm-driven economy.Cambridge: Harvard University Press, 2016, p.39)17

competitors’ pricing or activities9, there is no human order to execute, butthe use of an algorithm or the same algorithm by competitors that alignprices.Therefore, despite some legal challenges presented by those typesof collusion, they are not the subject of analysis of the present study sincethey do not represent a challenge to the antitrust toolkit.The focus of this paper is to discuss the other two scenarios:“predictable agent” and “autonomous machine” or “digital eye”. In those twocases the algorithms are unilaterally designed, in the end resulting in a tacitcollusion with no need of prior agreements among competitors. The“autonomous machine” is an even more grey zone, and it reflects thesituation in which the algorithm is designed exclusively to maximize profit.Moreover, the algorithm is complex enough to learn by itself and engage inan anticompetitive conduct – tacit collusion – with no influence of humans.In the “predictable agent” case, the algorithm is programmed tomonitor prices and increase profit, so when the price increase issustainable, the competitor follows and produces autonomous tacitcollusion10. The difference from the “autonomous machine” or “digital eye”scenario is the presence of the stability of the collusion since no deviationis possible without being caught by other competitors' algorithms and theabsence of any specific order to respond in a specific way to an rningandexperimentation11.In view of the enforcement challenge in those scenarios, this paperproposes the adoption of a mandatory compliance by design inprogramming such algorithms as an option to solve the matter.To achieve this task, in the first section of this article will be presentedthe typology of collusions on the digital markets, as well as the concepts9Ezrachi, Ariel and Stucke, Maurice E. Virtual Competition: the promise and perils of the algorithm-driveneconomy. Cambridge: Harvard University Press, 2016, p. 46.10 Ezrachi, Ariel and Stucke, Maurice E. Virtual Competition: the promise and perils of the algorithmdriven economy. Cambridge: Harvard University Press, 2016, p. 61.11 Ezrachi, Ariel and Stucke, Maurice E. Virtual Competition: the promise and perils of the algorithmdriven economy. Cambridge: Harvard University Press, 2016, p. 74.18

and functionalities of algorithms as means or reasons to facilitate,implement or result in a collusive behavior. The next section will evaluatethe efficacy of the traditional legal framework to combat collusion in theenvironment of the digital economy. The main problems, for instance theobstacles on evidencing authorship for collusive behavior derived from A.I.,will be discussed. The third section will explore possible solutions in thenormative level to address the problem of producing adequate antitrustenforcement of algorithmic collusion. Specifically, this section shall tacklethe idea of compliance by design, its potential to adjust the economicincentives for deterrence, vis-à-vis the legal hurdles for its concretion insome jurisdictions. Following the third section, the concluding remarks willbring up an organized presentation of the concepts, main reflections, andinsights from this innovative and rich discussion.Section I – Collusive behavior and algorithmsThe framework adopted for discussion is the theoretical definitionproposed by STUCKE & ERZACHI12 and already presented in theintroduction.The “messenger” and the “hub and spoke” cartel are well known bycompetition authorities and have the same effect than a hard-core cartel inany other market and several jurisdictions have already investigated andcondemned conducts of this model. Among the examples of “messenger”,it is worth mentioning the case of selling posters by Amazon by DavidTopkins13 which was investigated by the US Department of Justice and a12Although there are other classifications proposed by the specialized literature, the fact is that theframework proposed by Stucke and Erzachi was very well received by the academic environment, bothenabling a standard of background analysis with a

Diogo Kastrup Richter - LLM candidate in Economic Law and Development at Pontifícia Universidade Católica do Paraná (PUCPR/Brazil). Postgraduate diploma in Tax Law by the Instituto Brasileiro de Estudos Tributários (IBET/Brazil) and graduated in Law from PUCPR, with exchange program at Universidad de Zaragoza (UNIZAR/Spain). Lawyer.