EU ANTITRUST: HOT TOPICS & NEXT STEPS

Prague, Czechia

EU ANTITRUST: HOT TOPICS & NEXT STEPS 2022

A more aggressive approach to obtaining new competition authority powers and seeking to intrude beyond existing competition authority powers into algorithms may have a chilling effect and may hinder the development of technologies, including their undeniable positive impacts on the market and competition and ultimately on consumer welfare. Conventional wisdom suggests that if one doesn’t know (for sure) whether to do something, one simply shouldn’t do it. The Hippocratic maxim of “do no harm” should also apply in AT authorities’ responses to the still little explored phenomenon of pricing algorithms; the most important intervention may be the one that is not implemented. Unless regulatory humility prevails over regulatory paranoia, activism, and arrogance, the fear of throwing the baby out with the bathwater will be misplaced. Regulatory interventions should be limited to clear-cut cases or cases where there are overwhelmingly clear negative impacts (analogy to per se bans of the hardcore cartel type). Indeed, pricing algorithms can not only be a tool for cartelization, but also allow for a setup for price fighting with competitors, leading to lower prices and higher consumer welfare. Premature interventions in the possibilities of algorithms could lead to a somewhat schizophrenic situation (Kupčík, p. 544): undertakings should monitor the market in order to avoid price parallelism, which would indeed increase the transparency of the market, but which they would not be able to use because they are supposed to act independently. However, increased transparency can lead to unpunishable tacit collusion. A possible regulation of the frequency of price changes could in turn artificially freeze competition for prohibited periods. Thus, the focus can be partially freed from the “fixation on price fixing” and, thanks to perfect and previously unprecedented information, the focus can shift from price competition to differentiation of goods and services. “Competition can carry on, as a totally homogeneous market does not exist in reality” (Kupčík, ibid, p. 545). On the other hand, this radical proposal could lead to the legitimisation of cartel agreements under the pretext of using alleged monitoring algorithms. The law should not be construed as to equate the collusive equilibrium with a cartel violation (Thomas, p. 75). The possibility of prosecuting price fixing, regardless of the form in which it takes, must be maintained despite the increased evidentiary difficulties. Such conduct should be attributable to the operator running the algorithm, which has the potential to infringe Article 101 TFEU (Bejček 2020b, p. 391; Sonderegger, p. 225). However, it is hardly possible to distinguish between collusion caused by algorithms and collusion due to human action (as Sonderegger does; ibid, p. 225). I am convinced that in fact, even the forbidden agreement resulting from the algorithm is only a mediated human action consciously postponed and shifted to the robot in order to hide its illegal essence. The letter acrobatics of

355