PANEL: Corporate Accountability in AI Supply Chains: the AI-Workers' Rights Litigation (AI-WORLd)

Corporate accountability for workers' rights within AI supply chains remains deeply fragmented – or rather, systematically elusive. While lead firms formally externalize activities such as data labelling and content moderation through platforms and third-party contractors, operational control remains highly centralized through productivity metrics, data governance protocols, quality standards, and algorithmic surveillance. The fragmentation of AI production processes is not new: it reproduces a pattern familiar from the earliest industrial production systems – one in which workers are exploited, their rights systematically violated, and the legal liability of the companies who extract value deliberately obscured.

Yet the structural risks facing data workers who train, and feed AI systems have so far received limited attention from policymakers and regulators. Even the EU AI Act – widely presented as the most comprehensive regulatory framework addressing AI-related risks – focuses primarily on harms associated with the use of AI systems, while largely rendering invisible the significant labour and human rights risks borne by the AI workforce itself.

The AI-Workers' Rights Litigation (AI-WORLd) initiative seeks to challenge this frame. Through multiple strategic litigation initiatives targeting lead firms operating in the AI sector – relying in particular on corporate sustainability due diligence frameworks and data protection law – the project aims to hold leading AI services companies accountable for labour and human rights risks across AI supply chains.

By bringing together the project's multi-stakeholder partners of the project (trade unions, NGOs, legal practitioners, academics, and data workers' organisations), the panel will present the main legal challenges raised by the deployment of due diligence grounded strategic litigation for labour and human rights violations in AI supply chains. It will also explore the role of collective mobilisations and transnational alliances in shaping emerging forms of AI governance. On this terrain, the panel will analyse how strategic litigation may function – in a broader context marked by deregulation and growing political and cultural alignment with large technology companies – not only as a mechanism of legal enforcement, but also as a space for resistance, coalition-building, knowledge production, and political visibility for actors who remain largely invisible within dominant narratives of AI innovation.

Panel convenors: Baptiste Delmas (Sorbonne Law School, Paris, France) and Luca d'Ambrosio (Lysias Partners and University Paris 1 panthéon Sorbonne, France)

Panellists:

• Rachel Beausejour et Samira Alaoui (CGT-FAPT – France)

• Joan Muthoni Kinyua and Ephantus Kanyugi (Data Labelers Association – Kenya)

• Kimani Anthony Njenga (Oversight Lab Africa – Kenya)

• Anonymized (Content Moderator – Nigeria)