Law-abiding algorithms: How Big Data and AI challenges the competitive system
(2019) LAGF03 20192Department of Law
Faculty of Law
- Abstract (Swedish)
- I denna uppsats analyseras prissättningsalgoritmer för att undersöka hur de kan skapa nya sätt för företag att ägna sig åt kartellverksamhet. För att svara på frågan hur nuvarande konkurrensrättslig lagstiftning i EU kan appliceras på algoritmisk kartellbildning, studeras definitionen av de tre formerna av karteller i art. 101(1) FEUF. Genom ett studium av rättsfall, beslut från Kommissionen, samt litteratur på området kunde dessa definitioner appliceras på fyra olika sorters algoritmisk kartellbildning, the Messenger scenario, the Hub & Spoke scenario, the Predictable Agent scenario och the Digital Eye scenario. Resultatet av studien visar att kartellverksamhet genom prissättningsalgoritmer är svåra att lagföra under nu gällande... (More)
- I denna uppsats analyseras prissättningsalgoritmer för att undersöka hur de kan skapa nya sätt för företag att ägna sig åt kartellverksamhet. För att svara på frågan hur nuvarande konkurrensrättslig lagstiftning i EU kan appliceras på algoritmisk kartellbildning, studeras definitionen av de tre formerna av karteller i art. 101(1) FEUF. Genom ett studium av rättsfall, beslut från Kommissionen, samt litteratur på området kunde dessa definitioner appliceras på fyra olika sorters algoritmisk kartellbildning, the Messenger scenario, the Hub & Spoke scenario, the Predictable Agent scenario och the Digital Eye scenario. Resultatet av studien visar att kartellverksamhet genom prissättningsalgoritmer är svåra att lagföra under nu gällande konkurrensrättslig lagstiftning. Ett undantag är vid the Messenger scenario. Frånvaro av mänsklig inblandning i affärsbeslut och avsaknad av avtal eller överenskommelser mellan företag som ägnar sig åt kartellverksamhet resulterar i att förbudsregeln i art. 101(1) FEUF är svårt att tillämpa i dessa sitationer. Implikationen av dessa resultat är att myndigheter inte kan förbjuda kartelliknande utfall på marknaden när de aktuella företagen delegerar sitt beslutsfattande till en algoritm. (Less)
- Abstract (Swedish)
- In this thesis, the usage of pricing algorithms is analyzed to see how they facilitate new ways for companies to collude. To answer the question of how current competition legislation in the EU can be applied to algorithmic collusion, I studied the definition of the three forms of collusion in art. 101(1) TFEU through case law, Commission decisions, and literature on the subject. Subsequently, these definitions were applied to four scenarios of algorithmic collusion, the Messenger scenario, the Hub & Spoke scenario, the Predictable Agent scenario, and the Digital Eye scenario. The results of the study showed that collusion through algorithmic pricing models are difficult to enforce under current legislation in all scenarios but the... (More)
- In this thesis, the usage of pricing algorithms is analyzed to see how they facilitate new ways for companies to collude. To answer the question of how current competition legislation in the EU can be applied to algorithmic collusion, I studied the definition of the three forms of collusion in art. 101(1) TFEU through case law, Commission decisions, and literature on the subject. Subsequently, these definitions were applied to four scenarios of algorithmic collusion, the Messenger scenario, the Hub & Spoke scenario, the Predictable Agent scenario, and the Digital Eye scenario. The results of the study showed that collusion through algorithmic pricing models are difficult to enforce under current legislation in all scenarios but the Messenger scenario. The lack of human involvement in business decisions and the absence of an agreement between representatives of companies conducting collusive behavior renders the prohibitive rule in art. 101(1) TFEU difficult to apply. The implication of these findings is that the authorities cannot prohibit collusive outcomes on the market when the companies in question delegates decision-making to an algorithm. (Less)
Please use this url to cite or link to this publication:
http://lup.lub.lu.se/student-papers/record/9000108
- author
- Siöstedt, Anja LU
- supervisor
-
- Jacob Öberg LU
- organization
- course
- LAGF03 20192
- year
- 2019
- type
- M2 - Bachelor Degree
- subject
- keywords
- EU law, företagsekonomi, konkurrensrätt, competition law, law and economics, law, algorithms, algorithmic pricing, Artificial intelligence, Big Data, pricing algorithms, collusion, cartel, express collusion, tacit collusion
- language
- English
- id
- 9000108
- date added to LUP
- 2020-04-09 13:56:55
- date last changed
- 2020-04-09 13:56:55
@misc{9000108,
abstract = {{In this thesis, the usage of pricing algorithms is analyzed to see how they facilitate new ways for companies to collude. To answer the question of how current competition legislation in the EU can be applied to algorithmic collusion, I studied the definition of the three forms of collusion in art. 101(1) TFEU through case law, Commission decisions, and literature on the subject. Subsequently, these definitions were applied to four scenarios of algorithmic collusion, the Messenger scenario, the Hub & Spoke scenario, the Predictable Agent scenario, and the Digital Eye scenario. The results of the study showed that collusion through algorithmic pricing models are difficult to enforce under current legislation in all scenarios but the Messenger scenario. The lack of human involvement in business decisions and the absence of an agreement between representatives of companies conducting collusive behavior renders the prohibitive rule in art. 101(1) TFEU difficult to apply. The implication of these findings is that the authorities cannot prohibit collusive outcomes on the market when the companies in question delegates decision-making to an algorithm.}},
author = {{Siöstedt, Anja}},
language = {{eng}},
note = {{Student Paper}},
title = {{Law-abiding algorithms: How Big Data and AI challenges the competitive system}},
year = {{2019}},
}