MEPs adopt own-initiative report on AI in the workplace.

On 17 December, the plenary of the European Parliament adopted a legislative own-initiative report on digitalisation, artificial intelligence (AI) and algorithmic management in the workplace. The report calls on the European Commission to put forward rules on the use of algorithmic management at work. It was drafted by rapporteur Andrzej Buła (EPP, PL) in the Committee on Employment and Social Affairs (EMPL) and received broad cross-party support. The report was adopted by 451 votes to 45, with 153 abstentions.

Key considerations of the report

The report defines algorithmic management as automated monitoring and decision-making systems within the meaning of the Platform Work Directive. It notes that such systems are primarily used for core managerial and decision-making functions, including recruitment, work and task allocation, monitoring and control, performance evaluation, remuneration and dismissal decisions. MEPs stress that the use of such systems can offer significant opportunities, for example to optimise work processes, increase consistency and objectivity in management decisions, and improve occupational health and safety as well as flexibility at work, including through telework and flexible working time arrangements.

At the same time, the report highlights that algorithmic management and AI-based systems entail substantial risks. These include, in particular, increased time and performance pressure, work intensification, discrimination, intrusive monitoring and excessive control, bias effects in AI systems, stress and other psychosocial risks. While making use of the potential of these technologies, such risks must be prevented or adequately addressed. Against this background, the report sets out recommendations to the European Commission on the elements that should be included in a legislative proposal on algorithmic management in the workplace.

Recommendations for a legislative proposal

The report underlines that decisions taken or supported by algorithmic management must always be subject to effective human oversight. Workers should have the right to request a review where they consider that their rights have been infringed; as a result, the underlying system must be adapted or discontinued. According to the European Parliament, particularly sensitive employment-related decisions – such as recruitment, contract renewal or termination, remuneration and disciplinary measures – should be taken and reviewed exclusively by humans.

In addition, MEPs call for comprehensive transparency. Workers must be informed about when and how algorithmic management is used, which data it processes, how it affects their working conditions and how human oversight is ensured. Workers should also be consulted at an early stage when algorithmic management is used to support decisions on remuneration, performance evaluation, task allocation or working time. To protect workers’ privacy, the processing of particularly sensitive information should be prohibited, including data relating to emotional, psychological or neurological states, private communications, geolocation outside working hours, data use during non-working time, as well as data concerning freedom of association and collective representation.

Outlook

Following the adoption of the report by the plenary, the European Commission has three months to respond to the European Parliament’s request – either by informing Parliament of the planned next steps or by explaining why it does not intend to take action in line with Parliament’s demands. Executive Vice-President Roxana Mînzatu stated in the European Parliament that the Commission very much welcomes the report. On 4 December, the Commission published its Quality Jobs Roadmap. In this context, a social partner consultation was launched in relation to a possible legislative initiative, which will also address AI and algorithmic management in the workplace. The Commission will take further steps only after the consultation has concluded at the end of January 2026, according to Mînzatu.