The Use of Artificial Intelligence in the Worforce –
Robo-Firing and Monitoring
Whilst the use of technology for workforce management is nothing novel, covering various functions including monitoring attendance and preparing payroll, there is now a new phenomenon that has the potential to significantly impact the workforce: the use of artificial intelligence.
This article will provide insight into the use of AI to manage and monitor the performance of employees, a concept often referred to as ”algorithmic management”.
What is Algorithmic Management?
Algorithmic management refers to automated monitoring and decision-making systems through which digital labour platforms control or supervise the assignment, performance, evaluation, ranking, review of, and other actions concerning, the work performed by people working through platforms.
Algorithmic management encompass a diverse set of technology tools and techniques designed for remote workforce management and assessment, based on data collection and surveillance of workers to enable automated or semi-automated decision-making. Algorithmic management can include real time response to data influencing management decisions, and the use of ‘nudges’ and penalties to provide indirect incentivise for worker behaviour.
When this technology is used in the context of employee dismissals, a new term has emerged – “robofiring”. This term refers to the practice of using automated systems, algorithms, or AI tools to make decisions about employee retention or termination without direct human intervention. While these technologies promise objectivity and accuracy, they may inadvertently introduce bias and overlook vital contextual factors that humans consider.
The reach of these technologies extends beyond the gig economy, since they are now more affordable and easier to deploy. Such technology is increasingly becoming part of ordinary infrastructure of workplaces across different sectors.
Employment law implications
Employment laws are primarily designed to safeguard employee rights, prevent discrimination and ensure fair treatment at workplace. The specific challenges of algorithmic management in the platform work context are not covered by existing labour laws. However the adoption of these technologies is giving rise to a number of legal challenges concerning workers’ rights and may destabilise employment relationships due to:
- Bias and Discrimination: Automated systems might inadvertently rely on discriminatory patterns present in historical data, leading to biased decisions based on age, gender, race or other discriminatory factors.
- Transparency and Accountability: with complex algorithms, it can be challenging to access how the systems work, explain the rationale behind a robo fiing and potentially, hindering legal recourse for affected employees.
- Surveillance and control: Technology-enabled surveillance can generate new speed and efficiency pressures for workers and may lock workers out from important aspects of decision making, such as being able to use personal discretion.
Some of the legislative measures aimed to regulate specific aspects of this issue are briefly assessed hereunder.
EU’s proposed rules on improving the working conditions of platform workers
The proposed EU Platform Workers Directive, a legislative measure intended to address a number of issues relative to platform workers, includes specific rules for the use of artificial intelligence in the workplace.
The proposal identifies the issues concerning the digital control exercised by platforms (use of algorithms to assign tasks, but also to monitor, supervise, evaluate, impose sanctions and terminate the contract). The proposed directive would also increase the transparency in the use of algorithms by digital labour platforms, by introducing the requirement for human monitoring, in order to ensure fairness and accountability in algorithmic management and the respect of working conditions. Those performing human monitoring would also have the right to contest automated decisions. These new rights would be granted to both workers and genuine self-employed people.[1]
US state of Pennsylvania: No Robot Bosses Act
In the US state of Pennsylvania, a new law has been proposed under the name ‘No Robot Bosses Act’, which would forbid employers from making employment decisions solely based on automated systems, such as throughalgorithms and machine learning tools. Additionally, it would mandate that employers train their employees on how to use such systems and make sure employers disclose when they are using them.
The bill would also force employers to provide real-life, human oversight over AI’s use in the workplace before an automated system makes an employment-related decision — such as during the hiring process, when adding or removing work shifts, or when firing a worker based on their performance.
UK proposed Data Protection and Digital Information (No. 2) Bill
In the UK, the Data Protection and Digital Information (No. 2) bill has been proposed intendending to make data protection legislation simpler for businesses to understand and implement, has also been widely criticised due to the possible effect of reducing protection for workers against automated decision making afforded to them under Article 22 of the GDPR. The bill provides clarifications in relation to prohibition to automated decision making with no ‘meaningful human involvement’. The bill further states that when considering whether there is meaningful human involvement in the taking of a decision, a person must consider, among other things, the extent to which the decision is reached by means of profiling. The bill appear to weaken the right to require app-based companies to provide an explanation when they make automated decisions.
The UBER Case
Though relatively nascent, these concerns have already undergone judicial scrutiny in a highly significant case.
In April 2023, the Amsterdam Court of Appeal upheld the appeal filed by a group of drivers against the ride-hailing companies Uber and Ola Cabs. The drivers brought claims for breaches of the EU General Data Protection Regulation (“GDPR”) concerning decisions made using opaque algorithms which included managing, fining and sacking workers. Significantly, the Court of Appeal overturned the first instance ruling on this issue and rejected Uber's attempt to rely on the "humans in the loop" who were supposed to have reviewed and verified the algorithms' decisions. According to the Court, these reviews amounted to "not... much more than a purely symbolic act" in the context of the facts. As a result, the algorithmic decision-making was "solely automated”.
The Court also delved into the complex interplay between these rights and the protection of trade secrets and determined that the defendants were not permitted to refuse to disclose this information just because it constituted proprietary trade secrets.
Conclusion
The inevitable rise of AI and comparable technologies, coupled with their deployment into employment-related affairs, is poised to attract increased scrutiny.
These issues are poised to play a pivotal role in forthcoming court cases, where these matters will be further refined and elaborated.
Companies must take proactive measures to embrace automation while upholding employee rights, such as (i) auditing algorithms to ensure compliance with employment laws and detection of any potential bias; (ii) incorporating human oversight to the automated decision-making process; (iii) clearly communicating the use of automation in employment decisions; (iv) analyzing data collection to ensure workforce diversity and inclusion; and (v) consulting an employment law expert for legal advice.
[1] Briefing by European Parliamentary Research services dated June 2023 https://www.europarl.europa.eu/RegData/etudes/BRIE/2022/698923/EPRS_BRI(2022)698923_EN.pdf