Data sharing to address forced labour already takes place across sectors, but most efforts remain limited in scope and siloed by organization, geography or purpose. Confidentiality obligations, differing legal frameworks and a lack of trusted connective infrastructure restrict how far information can flow and how effectively it can be used. As a result, many stakeholders see only part of the overall picture. Federated data systems offer a way to connect these existing efforts safely. They enable information to be analysed where it resides, allowing participants to share insights without giving up control or sovereignty. Rather than creating another central database, the Partnership links existing ones through shared standards and secure protocols, turning isolated initiatives into a connected network. AI enhances this potential. Integrated throughout the federated system, it can identify patterns and relationships that individual datasets cannot reveal – linking, for example, worker grievances with inspection outcomes or migration flows – while protecting privacy through privacy- preserving computation. Together, these technologies make possible a form of collective intelligence at scale: a system where trust, accountability and analytical power reinforce one another.2.3 Why federated data and agentic AI are game changers Key drivers for effective data federation in the forced labour ecosystem BOX 6 –Regulations, investor expectations and supply chain due-diligence requirements are driving new demand for credible, connected data.20,21 Governments and companies now share a mutual interest in transparency, accountability and risk reduction. –Recent cross-sector partnerships have shown that sharing data in a precompetitive space allows organizations to manage systemic risks collectively while maintaining commercial independence.22 Connective infrastructure can scale these gains across industries and regions. –Privacy-preserving technologies and federated learning are now proven and practical, enabling collaboration without compromising confidentiality.23 Advances in AI and multilingual analysis make it possible to extract reliable insight from complex, unstructured information while maintaining data protection and sovereignty.24 –Computational capacity, cloud infrastructure and interoperability standards have matured, allowing distributed analytics at scale. International frameworks on data governance and responsible AI can provide guardrails for ethical adoption, creating conditions for innovation, measurable impact and shared accountability. Together, these advances align privacy, sovereignty and collaboration, demonstrating that secure data systems can deliver trust, efficiency and insight without centralizing information. Trust is built both technically, as each participant remains accountable for how they use their own and others’ information, and institutionally, through clear mutual agreements that define the rules of participation and information sharing. The following sections outline how this architecture operates in practice, how it differs from traditional data sharing models and why advances in privacy, interoperability and multilingual AI make this a pivotal moment for collaboration against forced labour. Harnessing Data and Intelligence for Collective Advantage: Ending Forced Labour in Global Supply Chains 13

Harnessing Data and Intelligence for Collective Advantage 2026