The diversity of forced labour data (Tech Against Trafficking’s Three Universes of Data) BOX 3 Forced labour data today spans three broad universes (corporate, civil society and public- sector systems) as described by Tech Against Trafficking’s Three Universes of Data. The challenge is not scarcity but fragmentation: connecting these universes securely is what the Partnership seeks to achieve. Universe: Corporate sector Examples of data types: Social audit reports, supplier compliance data, recruitment agency records, grievance data Typical sources: Companies, auditors, supply chain platforms Relevance to impact: Offers operational visibility and compliance evidenceUniverse: Civil society Examples of data types: Worker surveys, hotline data, union reports, NGO case files, survivor testimonies Typical sources: NGOs, unions, worker voice platforms Relevance to impact: Reveals lived experience, risk patterns and hidden coercion Universe: Public sector Examples of data types: Labour inspections, prosecutions, migration and border data, humanitarian assessments Typical sources: Government ministries, enforcement agencies Relevance to impact: Anchors prevalence data and supports policy design Source: Tech Against Trafficking. (2024). The current landscape of data sharing. Building an effective data ecosystem to address forced labor in global supply chains, pp. 24–32. This fragmentation is reinforced by deeper structural barriers that shape how data, incentives, trust and governance interact. Information about forced labour is abundant in some areas yet absent in others, especially in the “first mile” of production or informal work, where conditions remain largely invisible and under-documented. Each actor gathers data within their own mandate, using different tools, standards and incentives. As such, existing sources vary widely in quality and rarely align on standards or interoperability, while workers often under-report exploitation due to fear of retaliation or lack of access to reliable mechanisms.18 It is worth noting that differences in collection methods, verification standards and reporting incentives can produce misleading or incomplete information. Some datasets may reflect commercial, political or methodological bias, underscoring the importance of validation and triangulation across multiple sources before drawing conclusions.At the same time, incentives to share or even collect data are, in many cases, weak. When the benefits of transparency are uncertain, the risks high or exposure uncomfortable, stakeholders hesitate to disclose information. Uneven incentives and limited capacity for disclosure allow information gaps to persist across the system, particularly where commercial confidentiality, legal mandates or limited technical infrastructure constrain engagement. Compounding these barriers are persistent deficits of trust and governance. Privacy, sovereignty and reputational concerns deter collaboration, while longstanding mistrust between sectors reinforces silos and limits data exchange, even where objectives align. No shared governance framework exists to align accountability or measure collective impact, and information tends to flow vertically within sectors rather than horizontally across them. Harnessing Data and Intelligence for Collective Advantage: Ending Forced Labour in Global Supply Chains 7

Harnessing Data and Intelligence for Collective Advantage 2026