Contemporary discourse on artificial intelligence and work frames AI as a tool of managerial control, surveillance, and assessment.
This contribution aims to push the boundaries of current debate beyond this framework. Workers are not merely subjects of AI systems; they are, indeed, to be considered also their co-producers.
Through the generation of behavioural, corrective, and implicit feedback data in the ordinary course of employment, workers collectively train and improve the machine learning systems that govern AI and algorithmic management.
This productive contribution is commercially valuable, however it is also legally unrecognised, and entirely uncompensated. This happens also to the extent that the core obligations underpinning the employment contract are impacted by such technologies.
Drawing on the materialist tradition in labour studies, the political economy of digital labour, and the doctrinal resources of labour law and unjust enrichment, this speech develops the argument that workers who train AI systems are entitled to compensation for this additional, unpaid productive activity. It proposes a regulatory framework grounded in transparency, collective bargaining, and statutory minimum standards.
