Transparency in the datafied workplace: Law, workers, and job applicant perspectives

As AI systems become increasingly embedded in the workplace, and the datafication of recruitment and employment grows more pervasive, ensuring the trustworthiness of these technologies from design to deployment has emerged as a pressing concern - particularly within the European Union's evolving regulatory framework. The General Data Protection Regulation (GDPR) and the AI Act play a key role in shaping the governance of AI, including in recruitment and employment. Drawing on a large-scale survey conducted as part of the BIAS project, our study analyses 4317 valid responses from a diverse sample across the European Union, Iceland, Norway, Switzerland, and Turkey. It examines how job applicants and workers perceive and experience AI systems, their awareness of AI use, and their experiences with data processing and transparency mechanisms, paying particular attention to demographic patterns related to gender, age, and education. Our findings reveal that transparency measures, while sometimes present, remain fragmented and procedural. Opt-out mechanisms are also frequently unclear. AI systems in recruitment contexts collect substantially more data, particularly on sensitive topics. In workplace settings, data collection is lower and differences between sensitive and non-sensitive information are less pronounced. Subjective minority identification appears to be the most consistent demographic predictor of higher likelihood of data collection. Building on these findings, the article critically assesses how transparency is, and should be, operationalized under the GDPR and the AI Act. It argues for aligning legal compliance with practices that centre the experiences of those most affected and address structural inequalities in recruitment and employment.