Let’s start with why. Why was this project so urgent at the time?
Oliver: Everyone agrees that new technology in the workplace is fundamentally changing the world of work. In addition to the positive impacts, it also brings new challenges, such as assisting with tasks such as scheduling shifts and handing off repetitive tasks, and creating more space for creative tasks. These include workplace surveillance, increased pressure, and less control over data and privacy.
This project shows that collective bargaining is becoming an essential tool for regulating innovation in the workplace. At the same time, there is still much work to be done, as these topics have not yet been widely addressed in many countries. Important issues such as the right to object to automated decision-making, training staff and management on new AI tools, and the right to consult external data experts are often still left unwritten in contracts.
This project aims to raise awareness about the need to include issues related to AI and algorithmic management in collective bargaining. We also want to encourage discussions between trade unions and at the policy level. At the same time, we want to highlight good examples of agreements that can encourage trade unions to start negotiating on these topics.
Birte: When we launched this project in 2022-2023, AI was just emerging as a collective bargaining issue. Unions were inundated with concerns about health and safety, working time measurement and increased workloads, but there were significant gaps in knowledge. Workers and unions didn’t understand how these technologies worked, much less how to negotiate around them. UNI’s first survey of European member states in 2023 revealed a lack of information, consultation and awareness, even at a time when EU AI legislation was being debated. We realized that if we did not act, trade unions would remain ineffective in responding to and shaping this change.
Our first wave of research was a turning point. This study really showed that AI was and is a labor rights issue, not just a technology issue. And while the response has been remarkable, it’s still quite limited, with many of the expected countries participating, including our affiliates in the Nordic countries, Germany, and elsewhere. By the second wave, there were clear changes. Collective agreements containing AI clauses began to appear across Europe, particularly in Spain and Italy. We weren’t expecting it, but upon closer inspection, a few things became clear. Spain, for example, negotiated a landmark national framework that requires all new CBAs to include AI or digitalization clauses, paving the way for a series of new agreements that include references to AI and machine learning. Italy, on the other hand, has seen a huge upsurge, which can be explained not only by realism but also by the great interest of many trade unions who have made this topic a priority. Italy’s new agreement is incredibly pragmatic and clear, focusing on specific technologies such as AI agents in contact centers, rather than just vague statements. For example, instead of broad phrases like “workers need to be consulted,” name the technology like AI chatbots or algorithmic scheduling and set clear rules in negotiations. This specificity makes it easier to predict risks and negotiate safety nets.
How does this project fill a gap that cannot be addressed by regulation alone?
Oliver: Regulations such as the EU AI Act and Europe’s GDPR, which focus on data privacy and worker protection, as well as the EU Platform Work Directive, which includes an interesting chapter on algorithmic management, are important, but they are moving slowly and implementation is even slower. Collective bargaining, on the other hand, can act quickly at the local level and fill in the gaps that the law leaves behind. And these processes must occur simultaneously. Our project highlights that while some countries, such as Italy and Spain, are leading the way in the next wave of research (to be conducted in 2025), others are yet to catch up.
Birte: We’re not just collecting and analyzing CBA, we’re serious about incorporating what we learn into union advocacy, awareness, capacity building, and securing guidelines. However, this is not the case everywhere, and the divide between Central and Eastern Europe and Southern Europe still exists. This is because organizations with many day-to-day issues, ongoing strikes, and survival concerns do not have time for awareness training. It’s also a question of how comfortable you are with the situation in your country and whether you have the capacity to deal with the problem or whether other things take priority. Priorities can vary from country to country, and what’s more, why peer learning is so important is an important topic that should always be on the agenda. Technology companies don’t stop at national borders, so our projects share stories all over the world. We have seen agreements in Argentina and South Africa, and our Asian colleagues are eager to learn. AI is a global challenge and trade unions’ response must be as well.
Because of the interest in this topic, research travels far and wide and is presented at various high-level conferences across Europe. How can lessons learned from research be incorporated into policy agendas and debates?
Oliver: Over the past few years, we have held a significant number of workshops and presented our research at major conferences where the results of our research (Phase I and Phase II) were highlighted. Moving forward through collective bargaining In this series, we brought together trade unions, policymakers, and researchers to delve into the practicalities of AI negotiations. We introduced the dashboard, discussed the findings, and brainstormed ways to scale these practices across Europe. The key takeaway is that enforcement is just as important as the rules themselves. You can have the best CBA on AI, but if management ignores it, it’s not worth it. We are asking for resources and support for works councils to monitor compliance.
At Friedrich Ebert Stiftung’s competence center “Future of Work” in Brussels, we focus on the digital transformation of the world of work. We are working towards a fair future of work, where worker rights and good working conditions are as important as competitiveness and innovation. The regulation of new technologies in the workplace is therefore an important topic not only for trade unions but also for policy makers and civil society.
As Europe debates digital sovereignty, competitiveness and innovation, quality jobs and good working conditions are becoming increasingly important and real assets. To achieve this, it is important to use algorithmic management and AI fairly in the workplace. That’s why agreements and clear rules on these issues are so important. In some countries, the debate about regulating AI in the workplace is more advanced. Our aim is to promote this discussion in as many European countries as possible. That’s why we work with national offices across Europe to bring these discussions to national trade unions and policy makers. For example, we will hold another event on this project in France in April 2026 and will soon continue this work in Ireland in June 2026.
Some of the findings from conversations and interviews with union leaders include the risks of AI replacing workers or increasing workloads. What is the reality on the ground?
Birte: The truth is, no one yet fully understands its impact. Companies buy off-the-shelf AI tools without asking whether they are useful, legitimate, or even safe. For example, in contact centers, AI and algorithmic management can create a pressure cooker. Productivity increases, but so does burnout, stress, and health risks.
The Italian agreement shows the way forward. By naming the technology and its risks, unions can negotiate specific safeguards such as limits on surveillance, transparency in decision-making, and the right to human oversight. It’s about redistributing not only the burdens of AI but also its benefits.
What’s the next step for this project? How can we keep this momentum going?
Oliver and Birte: We focus on three interrelated areas:
- legal framework: Promote strengthening of existing regulations (such as GDPR) and new protections under AI law.
- consciousness: More events, more training, Labor.A 2024 Conference (Berlin)has presented his research to audiences around the world.
- Tools for trade unions: of dashboard and database is just the beginning.
Insights and the CBA database can help unions create checklists, find inspiration for model clauses, and tackle negotiation guides to make topics easier to understand and act on.
And we are looking to new horizons. The digital supply chain (content moderation, data annotation) is a hotspot. Workers in places like Kenya and the Philippines are organizing around the health and safety risks of their work in global AI supply chains. International solidarity is non-negotiable. Big Tech is global. So should our reactions.
Finally, if you are a trade union leader reading this, what should you do tomorrow?
Oliver: Start a conversation. look at us dashboard. Pick one clause and bring it to the next negotiation session. Even the smallest step can change the balance of power.
Birte: And we demand transparency. If your employer uses AI, ask, “What is AI?” How does it work? How does it impact employees? And is this really necessary for business? What justifies working with or using AI in relation to the work that needs to be done? Information is power. our 2024 survey and Project page Packed with examples that inspire action.
See more:
