Bereich: Arbeit & Chancengleichheit

This report is an updated version of the 2022 paper by Ursula Holtgrewe, Sander Junte and Barbara Glinsner. Since EU legislation on AI and algorithmic management was very much in flux at the time, the authors felt that an update would add value. For this version, Leonie Dworsky (ZSI) provided further insight into platform work and the Platform Directive. The report covers the development of AI and platform regulation, the AI Act and the Platform Directive up until April 2024, explores the positions of social partners and civil society on AI and work, and the outcomes of European social dialogue on the subject. 

We conclude that European regulation of AI and AM remain patchy. The European AI strategy and the AI Act prioritise market creation and regulation for AI over workers’ rights. Much political rhetoric frames the issues of work and employment in terms of a fundamental trade-off between innovation and regulation in which a “balance” must be found. In this context, job creation, skills and employability, and non-discrimination are the most market-compatible employment issues that orient policy. The Platform Directive is a more targeted attempt to regulate employment and work with some clarifications of employment status and workers’ rights.

Social partner initiatives in the private sector mirror the polarisation of digital-intensive labour markets. The joint declarations have been concluded only in the traditionally data-intensive telecommunications and insurance sectors, and a joint position achieved in the metal engineering sector. Umbrella organisations have concluded a process-oriented framework agreement on digitalisation. Employer associations and unions jointly aim for win-win configurations with skill provision, AI enhancing, not replacing human abilities, prevention and mitigation of discrimination and unfair bias. They go beyond the AI Act in also foreseeing somewhat clearer procedures of complaint and redress.

Yet behind the rhetoric of “balances” between competitiveness and workers’ rights lie very likely conflicts around information asymmetries, uncertainties over possibilities and practices of control and monitoring, and the transparency of algorithm-based decisions. These may well present quite fundamental challenges to social dialogue in the authors’ view since mutual trust and shared definitions of the situation are essential to any kind of bargaining.

It is the labour side of unions, labour and internet law experts and civil society organisations who favour a more “systemic” and dynamic approach to the governance of AI and AM with regard to the world of work. This may well require a distinct body of legislation that gives human, social and workers’ rights more of a priority and bring AI governance up to the standard of the more ambitious legislation on worker rights of information and representation, anti-discrimination and data protection.

The report can be downloaded HERE.

For firsthand insights into social partners’ own views on the practice and regulation of AI and AM in Europe and their organisations’ strategies, also see the related paper: 

AutorInnen: Dworsky, L., Glinsner, B., Holtgrewe, U., Sander Junte (QUIT)

Verwandte Artikel:

• Projekt: Democracy at Work through Transparent and Inclusive Algorithmic Management

Tags: artificial intelligence, Europe, social dialogue