Emotion Recognition Systems under the AI Act: Legal, Technological, Societal and Psychological Perspectives
Emotion Recognition Systems under the AI Act: Legal, Technological, Societal and Psychological Perspectives
Emotion recognition systems and fundamental rights
Emotions play a defining role for social interactions, learning, and decision-making. Today, individuals are increasingly confronted with AI systems that attempt to recognise emotions through various approaches such as affective computing techniques. These systems are increasingly used and developed despite widespread criticism, their shaky assumptions, reliability, and trustworthiness. The use of AI systems that attempt to infer emotions or intentions creates risks for fundamental rights, public values, and society at large.
The AI Act takes a risk-based approach to regulate AI systems posing risks to fundamental rights, health and safety of individuals. For this reason, it also specifically regulates 'emotion recognition systems': AI systems aiming to infer emotions or intentions of natural persons based on their biometric data. The AI Act prohibits the placing on the market, putting into service or the use of AI systems to infer emotions at the workplace and in education. Outside these contexts, the AI Act classifies emotion recognition systems as high risk, meaning they are subject to specific product requirements.
A multidisciplinary perspective
This technology and its regulation touches on several disciplines including Affective Computing, although their approaches and key points of concern differ. A multi-disciplinary approach is needed to better understand and effectively mitigate the risks posed by this technology — including discrimination and the curtailment of human autonomy and dignity for those subjected to it.
Last year, the Autoriteit Persoonsgegevens (AP) published the fifth edition of its Report AI & Algorithms Netherlands (RAN), with a special focus on the use and risks of emotion recognition systems. The Open Universiteit (OU) and the Autoriteit Persoonsgegevens (AP) organise this event to bring together the legal, technological, societal and psychological aspects of emotion recognition systems and to facilitate a cross-disciplinary and multi-stakeholder discussion on the accompanying challenges. This event will explore questions such as:
- Is it possible to infer and categorise emotions computationally?
- Which approaches are used to infer emotions through technical means?
- For which applications are emotion recognition systems developed and used? What risks does this technology pose for individuals and society?
- How does the AI Act regulate emotion recognition systems? Does it sufficiently cover all use cases and mitigate the risks of this technology?
Programme
| 13.00 - 13.30 | Registration and coffee |
| 13.30 - 13.45 | Opening Representative of the Autoriteit Persoonsgegevens (AP) |
| 13.45 - 14.15 | Keynote Andrew McStay, PhD Professor of Technology & Society, Bangor University |
| 14.15 - 14.35 | Technological perspectives on emotion recognition Dr. ir. Joost Broekens Associate Professor Affective Computing and Human-Robot Interaction, Leiden University |
| 14.35 - 14.55 | Legal perspectives on emotion recognition and the AI Act Dr. Andreas Häuselmann LL.M. Assistant Professor of Privacy and Data Protection Law, Open Universiteit |
| 14.55 - 15.15 | Break |
| 15.15 - 15.35 | Psychological perspectives on emotion recognition Dr. Jessica Szczuka Research Center Trustworthy Data Science and Security at University of Duisburg-Essen |
| 15.35 - 16.30 | Panel discussion with all speakers and Q&A |
| 16.30 - 16.40 | Wrap-up and main takeaways |
| 16.40 - 17.30 | Drinks |
Practical information
Date: Wednesday 16 September 2026, 13.00 - 17.30
Location: University Campus Spui, The Hague
Registration
This event is open to students, researchers, policymakers, legal professionals, technology experts, and other professionals interested in the legal, technological, societal, and psychological aspects of emotion recognition systems.;
Please register by completing the registration form. Registration is free of charge, and capacity for this event is limited.
This event forms part of the Sector Plan Social Sciences and Humanities (SSH).
