Changes from plenary report to adopted text

Recommendation 1

Subject matter and scope

The aim of the proposal should be to improve working conditions and to promote the transparent, fair, accountable and safe deployment and use of automated monitoring and decision-making systems used to take or to support decisions at the workplace, ensuring human oversight, the protection of workers’ fundamental rights, and the protection of workers in relation to the processing of their personal data.

The proposal should apply to every worker and employer in the Union. Workers in a contractual relationship with an intermediary should enjoy the same level of protection as workers with a direct contractual relationship with their employer.

The proposal should respect the principle of proportionality and should ensure that the administrative and compliance burden imposed is appropriate to the size of the employer and the resources at its disposal, the nature of the technologies used, and the level of the risk involved, particularly with regard to micro, small and medium-sized enterprises.

Recommendation 2

Definitions

‘algorithmic management’ should be defined as automated monitoring systems and automated decision-making systems, in accordance with the Platform Work Directive..

The proposal should also provide definitions of: ‘worker’, ‘employer’, ‘worker’s representatives’ and ‘workplace’ in accordance with Union labour law including the case law of the Court of Justice, to avoid discrepancies.

Recommendation 3

Transparency and right to information

The proposal should ensure that employers provide the workers affected and the workers’ representatives concerned with information, in writing and in an accessible format, concerning the use or planned use of systems for algorithmic management at the workplace.

Such information should be provided in accordance with the General Data Protection Regulation and should include:

(a) a clear statement that algorithmic management systems are in use or intend to be introduced to carry out automated or semi-automated decision-making, including a general description of the purpose and the decisions taken or supported by those systems, their functioning and the workers and processes affected by them, the nature of such information being proportional and limited to what is necessary for the workers to carry out their work, to understand how the algorithmic systems influence decisions affecting them, and to exercise their rights;

(b) the impact of such systems on the workers’ working conditions and employment status;

(c) the categories of data collected or otherwise processed by such systems in relation to the workers, the purpose of the processing of the data and the details of the persons to whom the data is transferred;

(d) the human oversight mechanisms;

(e) training and support measures for the workers directly affected.

The information should be provided at the latest on the first working day of the affected workers, prior to the introduction of changes substantially affecting their working conditions, and at any time upon the request of the workers or the workers’ representatives.

The information should be provided in a clear, transparent and easily understandable and accessible manner, using plain language, which is adapted to the level of digital literacy reasonably expected of the workers, and should be accessible to persons with disabilities. The competent authorities and the workers’ representatives should have the right to ask for additional clarifications and details regarding any of the information provided.

Candidates undergoing a recruitment procedure should receive relevant information on automated decision-making systems used during the recruitment procedure at the time they apply for a vacancy. Workers should be provided with tools for the effective exercise of their data access rights and of their portability rights.

Recommendation 4

Consultation

The proposal should ensure that the deployment and use of new systems of algorithmic management, or updates to existing systems, which directly affect workers' remuneration, evaluation, working arrangements, task allocation or working time, are deemed to be decisions likely to lead to substantial changes in work organisation or in contractual relations and as such are subject to consultation pursuant to Article 4(2), point (c), of Directive 2002/14/EC, without prejudice to Directive 2009/38/EC.

Such consultation should be carried out with regard to:

(a) the objectives of the deployment or update, the work processes and the workers affected;

(b) the workload, work intensity, scheduling, working time, flexibility, renumeration, job content, performance evaluation or contractual relations;

(c) the effects on occupational health and safety;

(d) the types of data collected;

(e) the measures in place to prevent, detect, mitigate and correct bias or discriminatory outcomes;

(f) human oversight and review mechanisms;

(g) training and support measures, including the assistance of experts, for the workers and the managers deploying and using such systems.

Recommendation 5

Prohibited practices

The proposal should prohibit the collecting, storing, monitoring, sale or any other processing of:

(a) emotional, psychological or neurological states, cognitive activity or biometric data of the workers;

(b) private communications, including exchanges with other workers and workers’ representatives, except where such processing is lawful in accordance with Article 6(1), point (c) or (d) of the General Data Protection Regulation;

(c) data of workers while off-duty, not offering or performing work or in private spaces, real-time geolocation tracking outside working hours or beyond what is strictly necessary for the execution of tasks;

(d) data for the prediction of the exercise of fundamental rights, including the freedom of association, the right of collective bargaining and action or the right to information and consultation;

(e) special categories of personal data in accordance with Article 9(1) of the General Data Protection Regulation as well as inferences about such categories.

Those prohibited practices should also apply throughout the recruitment procedure.

Recommendation 6

Human oversight and review

The proposal should ensure continuous, meaningful and effective human oversight at all times over all decisions taken or supported by algorithmic management systems. It should also ensure that employers designate and inform workers and their representatives about the entities responsible for the monitoring of the functioning and impact of such systems and for reviewing their decisions. The persons responsible for oversight and evaluation should have the competence, training and authority necessary to exercise those functions, including the authority to override automated decisions.

The proposal should ensure that workers have the right to obtain from the employer, in reasonable time, a meaningful and comprehensible oral or written explanation regarding any decision taken or substantially supported by algorithmic management that affects the essential aspects of their employment, including the allocation of tasks, the evaluation of performance, working time scheduling, remuneration, and disciplinary measures, where any decision relating to such matters has been taken or has been substantially influenced by an algorithmic system. If the worker perceives his or her rights to be infringed by the decisions taken or substantially supported by algorithmic management systems, he or she should have the right to ask for review of that decision.

Where, as a result of human review, an individual decision taken or supported by algorithmic management is found to have infringed the rights of a worker, the employer should take the steps necessary, including, if appropriate, the modification of the algorithmic management system or the discontinuation of its use, and adequate compensation for the damage sustained.

Decisions concerning the initiation or termination of employment, the renewal or non-renewal of a contractual agreement, or any changes in remuneration or disciplinary action should always be taken by a human being and should be subject to human review.

The proposal should ensure that workers or their representatives may request the employer to trigger a review of the functioning of the algorithmic management systems in place where there is a justified concern that such systems exhibit systemic biases, discrimination or faults, or pose a threat to the mental or physical well-being of workers or to health and safety at the workplace.

Recommendation 7

Occupational health and safety

The proposal should ensure that employers integrate, in cooperation with workers’ representatives, the evaluation of the risks of algorithmic management systems into their safety and health systems, as regards possible risks of work-related accidents, psychosocial and ergonomic risks as well as undue pressure put on workers.

Employers should use algorithmic management systems in a manner that respects workers’ wellbeing and does not put at risk their safety or their physical and mental health. They should take mitigating actions where appropriate.

Recommendation 8

Responsibilities of competent national authorities

The proposal should ensure that labour inspectorates or other national competent authorities such as equality bodies, where appropriate, oversee the safe, transparent and non-discriminatory use of algorithmic management systems in the workplace.

Labour inspectorates should be tasked with monitoring, controlling and assessing the safety of algorithmic systems used in employment contexts, in particular with regard to the impact of those systems on the physical and mental health of workers and the fundamental rights of those workers.

Labour inspectorates should be provided with adequate resources, powers, technical expertise and appropriate training to effectively carry out their functions without duplicating existing controls or compliance mechanisms under national law. Upon request, they should have access to information on the use of algorithmic management in the workplaces.