Governing artificial intelligence-enabled labour surveillance: A multi-level framework for legal, organisational and collective governance in the digital workplace

Artificial intelligence (AI)-enabled labour surveillance marks a qualitative shift in industrial relations – less because it increases oversight than because it relocates control from observable conduct to inferred states, with knock-on effects for voice and accountability. This paper develops a multi-level framework linking legal regulation, organisational governance and collective representation to explain why existing industrial relations institutions struggle to absorb this shift. Drawing on labour process theory and comparative political economy, it argues that predictive scoring, affective inference and workplace datafication generate cross-level institutional misalignment rather than a simple regulatory gap. Through a cross-jurisdictional institutional comparison centred on the European Union AI Act, the General Data Protection Regulation and platform labour-related regulation, and compared with governance practices in the United States and selected Asian jurisdictions, the paper identifies a recurrent triad of failure: legal controls remain largely ex post and abstract; organisational policies are adaptable yet frequently unilateral, and collective bargaining and consultation face technical opacity and information asymmetries that limit substantive oversight. The paper's theoretical contribution is to reconceptualise AI labour surveillance as a problem of institutional translation. The paper conceptualises a ‘sustainable digital labour contract’ as a governance device that operationalises legal duties through participatory design and collectively negotiated protections for data access, grievance and algorithmic impacts.