Algorithmic management and the GDPR
(2022) JURM02 20221Department of Law
Faculty of Law
- Abstract
- The purpose of this thesis is to examine the effectiveness of the GDPR as a tool for protection for algorithmically managed workers in light of the standard of protection offered for working conditions in the EU. Algorithmic management refers to the use of automated means to direct, sanction and evaluate workers without human involvement, and it is an especially prevalent phenomenon in the platform economy. The use of algorithmic management challenges the conditions of work, as workers might not have any insight into what factors the automated decisions they are subject to are based on and there can be difficulties in challenging such decisions when they are erroneous or unfair. These challenges reinforce the power and information... (More)
- The purpose of this thesis is to examine the effectiveness of the GDPR as a tool for protection for algorithmically managed workers in light of the standard of protection offered for working conditions in the EU. Algorithmic management refers to the use of automated means to direct, sanction and evaluate workers without human involvement, and it is an especially prevalent phenomenon in the platform economy. The use of algorithmic management challenges the conditions of work, as workers might not have any insight into what factors the automated decisions they are subject to are based on and there can be difficulties in challenging such decisions when they are erroneous or unfair. These challenges reinforce the power and information asymmetries between workers and employers. Furthermore, these challenges are difficult to address due to regulatory challenges resulting from the inherent characteristics of AI systems – the technical complexity and lack of transparency - which can contribute to a lack of accountability. Working conditions are protected in the EU Charter of Fundamental Rights, in its Article 31(1), which states that every worker has the right to fair and just working conditions.
The thesis finds that the GDPR provisions on automated decision-making have shortcomings which affects the protection offered to workers when they are subject to automated decision-making. Article 22(1) stipulates a right to not be subject to solely automated decision-making which has a legal or similarly significant effect on the individual. This thesis highlights that it is firstly, unclear whether the provision constitutes a ‘right to oppose’ automated decision-making or a prohibition with some exceptions. Furthermore, it is unclear whether automated decisions taken in the context of employment – on the direction of workers, on the sanctioning and evaluation of workers – have ‘legal or similarly significant effects’ within the meaning of the Article.
The protection offered by the GDPR could be strengthened by a clarification of what types of automated decisions in the context of employment are strictly necessary for the performance of a contract and by clarification on what decisional outcomes have ‘legal or similarly significant effects’ within the meaning of Article 22(1). The rights to meaningful information on automated decision-making could be extended to all automated decision-making. A proposal for a directive with the aim of improving working conditions in platform work was published in December 2021, which provides a more robust right to information on all automated decisions which affect working conditions significantly. Furthermore, the proposed directive gives more insight for workers in how their behaviour is monitored and how their behaviour subsequently affects the automated decisions they are subjected to.
A uniform level of protection for workers could further be achieved by allowing all workers to enjoy the rights related to algorithmic management offered in the proposed platform work directive. (Less) - Abstract (Swedish)
- Syftet med denna uppsats är att undersöka effektiviteten av GDPR som ett verktyg för att skydda arbetstagare från negativa konsekvenser som kan uppstå i samband med algoritmisk arbetsledning. Algoritmisk arbetsledning är ett allmänt förekommande fenomen speciellt i plattformsekonomin, som avser användningen av automatiserade medel för att styra arbetstagare i utförandet av det dagliga arbetet. Automatiserade beslut inom plattformsekonomin kan tas om ersättning, om fördelningen av arbetsuppgifter och i utvärderingen av hur arbetet utförs. Arbetstagare kan ha bristande information om vilka faktorer som besluten grundar sig på och därmed bristande möjligheter att bestrida automatiserade beslut när de är felaktiga eller orättvisa. Dessa... (More)
- Syftet med denna uppsats är att undersöka effektiviteten av GDPR som ett verktyg för att skydda arbetstagare från negativa konsekvenser som kan uppstå i samband med algoritmisk arbetsledning. Algoritmisk arbetsledning är ett allmänt förekommande fenomen speciellt i plattformsekonomin, som avser användningen av automatiserade medel för att styra arbetstagare i utförandet av det dagliga arbetet. Automatiserade beslut inom plattformsekonomin kan tas om ersättning, om fördelningen av arbetsuppgifter och i utvärderingen av hur arbetet utförs. Arbetstagare kan ha bristande information om vilka faktorer som besluten grundar sig på och därmed bristande möjligheter att bestrida automatiserade beslut när de är felaktiga eller orättvisa. Dessa utmaningar förstärker informations- och maktasymmetrierna mellan arbetstagare och arbetsgivare. Teknologier som grundar sig på artificiell intelligens kännetecknas av bristande transparens och komplexitet som försvårar den effektiva regleringen av dessa teknologier.
GDPR reglerar användningen av automatiserat beslutsfattande och ger enskilda rätt till insyn och information vid sådant beslutsfattande. Arbetsvillkoren skyddas i EU:s stadga om de grundläggande rättigheterna, vars artikel 31(1) stadgar att varje arbetstagare har rätt till rättvisa arbetsförhållanden.
Artikel 22(1) GDPR föreskriver en rätt att inte blir föremål för ett beslut som enbart grundas på automatiserat beslutsfattande, vilket har rättsliga följder för honom eller henne eller på liknande sätt i betydande grand påverkar honom eller henne. Uppsatsen belyser hur bestämmelserna kring automatiserat beslutsfattande har brister som kan påverka det skydd som erbjuds arbetstagare när de är föremål för automatiserat beslutsfattande. Uppsatsen finner att det är oklart om bestämmelsen utgör en rätt för individer att motsätta sig automatiserat beslutsfattande, eller om det är fråga om ett förbud med vissa undantag. Det är vidare oklart om automatiserade beslut som fattas inom anställningen med syftet att leda, sanktionera eller utvärdera arbetstagare har rättsliga följder eller följder som på liknande sätt i betydande grad påverkar arbetstagaren. Det finns ett behov för tydlig vägledning i hur artikeln ska tolkas eftersom det endast finns en rätt till meningsfull information om automatiserat beslutsfattande om sådant beslutsfattande faller inom tillämpningsområdet för artikel 22(1). Ett förslag till ett direktiv om bättre arbetsvillkor för plattformsarbete publicerades i december 2021. Direktivet syftar till att förbättra arbetsvillkoren i för plattformsarbetare genom att introducera en mer djupgående rätt till insyn om automatiserat beslutsfattande som i betydande grad påverkar plattformsarbetstagarnas arbetsvillkor. Vidare avser direktivet att ge mer insikt till arbetstagare i hur deras utförande av arbetet övervakas och hur utvärderingen av utförandet av arbetet vidare påverkar det automatiserade beslutsfattandet. Direktivet är riktad till personer som utför arbete i plattformsekonomin, men ett mer enhetligt skydd för arbetstagare inom EU kan uppnås om bestämmelserna om algoritmisk arbetsledning i direktivet utvidgas till att omfatta arbetstagare utanför plattformsekonomin. (Less)
Please use this url to cite or link to this publication:
http://lup.lub.lu.se/student-papers/record/9080568
- author
- Tuomaila, Aleksandra LU
- supervisor
- organization
- course
- JURM02 20221
- year
- 2022
- type
- H3 - Professional qualifications (4 Years - )
- subject
- language
- English
- id
- 9080568
- date added to LUP
- 2022-06-12 15:34:50
- date last changed
- 2022-06-12 15:34:50
@misc{9080568,
abstract = {{The purpose of this thesis is to examine the effectiveness of the GDPR as a tool for protection for algorithmically managed workers in light of the standard of protection offered for working conditions in the EU. Algorithmic management refers to the use of automated means to direct, sanction and evaluate workers without human involvement, and it is an especially prevalent phenomenon in the platform economy. The use of algorithmic management challenges the conditions of work, as workers might not have any insight into what factors the automated decisions they are subject to are based on and there can be difficulties in challenging such decisions when they are erroneous or unfair. These challenges reinforce the power and information asymmetries between workers and employers. Furthermore, these challenges are difficult to address due to regulatory challenges resulting from the inherent characteristics of AI systems – the technical complexity and lack of transparency - which can contribute to a lack of accountability. Working conditions are protected in the EU Charter of Fundamental Rights, in its Article 31(1), which states that every worker has the right to fair and just working conditions.
The thesis finds that the GDPR provisions on automated decision-making have shortcomings which affects the protection offered to workers when they are subject to automated decision-making. Article 22(1) stipulates a right to not be subject to solely automated decision-making which has a legal or similarly significant effect on the individual. This thesis highlights that it is firstly, unclear whether the provision constitutes a ‘right to oppose’ automated decision-making or a prohibition with some exceptions. Furthermore, it is unclear whether automated decisions taken in the context of employment – on the direction of workers, on the sanctioning and evaluation of workers – have ‘legal or similarly significant effects’ within the meaning of the Article.
The protection offered by the GDPR could be strengthened by a clarification of what types of automated decisions in the context of employment are strictly necessary for the performance of a contract and by clarification on what decisional outcomes have ‘legal or similarly significant effects’ within the meaning of Article 22(1). The rights to meaningful information on automated decision-making could be extended to all automated decision-making. A proposal for a directive with the aim of improving working conditions in platform work was published in December 2021, which provides a more robust right to information on all automated decisions which affect working conditions significantly. Furthermore, the proposed directive gives more insight for workers in how their behaviour is monitored and how their behaviour subsequently affects the automated decisions they are subjected to.
A uniform level of protection for workers could further be achieved by allowing all workers to enjoy the rights related to algorithmic management offered in the proposed platform work directive.}},
author = {{Tuomaila, Aleksandra}},
language = {{eng}},
note = {{Student Paper}},
title = {{Algorithmic management and the GDPR}},
year = {{2022}},
}