How to identify forced labour in your supply chain

Identifying forced labour is one of the most challenging aspects of supply chain due diligence. It does not appear on standard audit checklists.

Workers experiencing coercion or exploitation rarely disclose it to people they have not met before. And the structural conditions that enable forced labour are often invisible unless you know specifically what to look for.

This page explains the established indicators of forced labour, why many conventional identification methods are poorly suited to detecting it, and what more effective approaches look like in practice.

For background on what forced labour and modern slavery mean, see: What is modern slavery? Definitions, forms and scale.

Why forced labour is difficult to detect

Before turning to the indicators and methods, it is worth being clear about why identification is hard. This shapes everything about how it should be approached.

Forced labour is designed to be hidden. Workers who are coerced or controlled have strong reasons not to disclose their situation to unknown visitors, particularly when those visitors are connected to – or operating with the knowledge of – their employer. The fear of losing income, being deported, having debts called in, or facing retaliation are all real deterrents to disclosure.

It often looks like ordinary employment. Workers in forced labour are frequently on payroll, receive some wages, and have signed contracts. The distinction between exploitative–but–legal conditions and forced labour lies in whether workers are genuinely free to leave – which is not something a document review or brief walkthrough of a worksite will reveal.

It concentrates where visibility is lowest. Forced labour in supply chains tends to occur in informal tiers – among subcontractors, homeworkers, labour agencies, and raw material producers – precisely where brands and retailers have least visibility and where audit programmes rarely reach.

Standard audits are not designed for it. A social audit scheduled in advance, conducted with the employer’s cooperation, and relying on worker interviews in the employer’s facility creates conditions in which workers experiencing coercion will almost certainly not speak honestly. The ILO and numerous independent researchers have documented this limitation consistently. Audit–based detection of forced labour is the exception rather than the rule.

This does not mean identification is impossible. It means it requires different approaches, combined with a realistic understanding of what each method can and cannot surface.

Human rights due diligence (HRDD): a complete guide for business

The ILO’s eleven indicators of forced labour

The most widely used framework for identifying forced labour is the set of eleven indicators published by the ILO. These were developed to help labour inspectors, auditors, and businesses identify when a situation of forced labour may be present. They are indicators, not a checklist: no single indicator is automatically conclusive, and context matters significantly. But where several are present, or where any single one is present in a severe form, the situation warrants serious investigation.

1. Abuse of vulnerability
Workers are targeted because of circumstances that make them easier to control: poverty, immigration status, language barriers, lack of social networks, or discrimination. Recruiters and employers exploit these vulnerabilities to create dependency. Warning signs include workforces composed overwhelmingly of migrant workers, workers housed in employer–controlled accommodation, or workers with limited ability to communicate in the local language and no knowledge of local legal rights.

2. Deception
Workers were not told the truth about the nature, location, terms, or conditions of the work they were recruited for. This may mean promised wages that differ from actual pay, a different job role than advertised, a different country of destination, or working conditions that are materially worse than represented. Deception in recruitment is one of the most common pathways into forced labour, particularly for migrant workers.

3. Restriction of movement
Workers are unable to move freely – physically confined to a worksite, dormitory, or vehicle, or prevented from leaving employment. Restriction of movement can be overt (locked gates, surveillance) or practical (workers housed in remote locations with no transport access, or told their documents will not be returned if they leave).

4. Isolation
Workers are kept away from the wider community, from other workers outside their employer’s control, or from anyone who might help them seek assistance. This may take the form of remote worksite locations, confiscation of mobile phones, or prohibition on contact with family, trade union representatives, or civil society organisations.

5. Physical and sexual violence
Workers are subjected to or threatened with violence, including physical assault, sexual abuse, or harm to family members. Physical violence is the most extreme and visible indicator, but it coexists with more subtle forms of control and may only be disclosed through trusted, confidential channels.

6. Intimidation and threats
Workers are threatened with consequences for non–compliance or attempts to leave: threats of violence, denunciation to immigration authorities, false criminal accusations, withdrawal of housing or food, or reputational harm to the worker or their family. Threats are often implied rather than explicit and may be difficult to document without building genuine trust with workers.

7. Retention of identity documents
Employers, labour recruiters, or supervisors hold workers’ passports, national identity cards, work permits, or other identity documents – ostensibly for safekeeping, but in practice removing the worker’s ability to leave independently or access services. This practice is particularly common in sectors relying on migrant labour and is explicitly prohibited by ETI’s Base Code (Clause 1) and international standards. No legitimate employment relationship requires workers to surrender their documents.

8. Withholding of wages
Wages are not paid at all, paid partially, paid irregularly, or subject to deductions that the worker has not consented to and cannot query. This includes deductions for accommodation, food, tools, or transport that were not disclosed at recruitment, excessive deductions for debt repayment, and wage advance arrangements that create inescapable financial dependency.

9. Debt bondage
Workers owe money to their employer or recruiter – most commonly as a result of recruitment fees paid to access employment – and find themselves unable to leave because the debt is larger than they can repay, grows faster than it can be cleared, or is subject to terms they never agreed to or did not understand. The ILO’s Fair Recruitment principles are unambiguous: workers should never pay fees to access employment. Where recruitment fees have been charged, debt bondage risk is high.

10. Abusive working and living conditions
Conditions in the workplace or accommodation are dangerous, degrading, or significantly below legal or contractual standards. This includes exposure to hazardous conditions without protection, extreme overcrowding in employer–provided housing, inadequate sanitation, or withholding of food. These conditions both indicate potential forced labour and create the vulnerability that enables other forms of control.

11. Excessive overtime
Workers are required to work hours significantly in excess of legal maximums or contractual terms, with no meaningful ability to refuse. This may occur alongside other indicators – workers in debt bondage or with restricted movement have no practical ability to decline excessive overtime – or may be a standalone indicator in situations where workers fear the consequences of refusal.

How the indicators interact in practice

The eleven indicators rarely appear in isolation. In practice, forced labour situations typically involve several operating together: a migrant worker who paid a recruitment fee (debt bondage) to take a job that was described differently from reality (deception), whose documents are held by their employer (retention of identity documents), who lives in employer–provided accommodation in a remote location (restriction of movement, isolation), and who is required to work excessive hours with deductions from their already–insufficient wages (excessive overtime, withholding of wages).

Understanding this interconnection matters for two reasons. First, it means that identifying one indicator should prompt investigation of others – the indicators are an entry point to a fuller picture, not a standalone finding. Second, it means that vulnerability and exploitation exist on a continuum: workers may experience some indicators without being in a situation that meets the legal threshold of forced labour, while being at significant risk of crossing that threshold if conditions worsen. Any indicators present warrant attention, not just those that clearly cross the line.

Where to look: risk mapping before investigation

Effective identification starts before any site visit or worker interview. A risk–based approach means directing investigative effort toward the parts of the supply chain where forced labour risk is highest.

Map your supply chain beyond tier one. Most forced labour in supply chains occurs not in the factories and farms that brands directly contract with, but in lower tiers – subcontractors, homeworkers, raw material producers, and labour agencies. Without visibility of these tiers, identification efforts will systematically miss the highest–risk populations.

Assess labour sourcing practices. How are workers recruited for your suppliers? Are recruitment agencies involved? Are fees charged to workers? Are there migrant workers in the workforce and, if so, from which countries and through which channels? Labour recruitment is one of the highest–risk points in a supply chain and one of the easiest to investigate through documentation and supplier questions.

Identify structural risk factors. Certain combinations of factors significantly elevate forced labour risk: reliance on migrant labour from high–risk corridors; seasonal production that creates labour demand spikes met by informal agencies; very low wages in the context of high cost of living; remote worksite locations; lack of worker representation or active discouragement of union membership.

Use external data sources. Civil society organisations, trade unions, investigative journalists, and government bodies (including the US Department of Labor’s List of Goods Produced by Child Labor or Forced Labor, and the UK’s Modern Slavery Act registry of statements) hold significant knowledge about forced labour risks in specific sectors and geographies. Engaging with these sources before investigation is more efficient and more reliable than relying solely on internal data.

Identification methods: what works and what doesn’t

Social audits: a limited but not worthless tool

Standard social audits – scheduled in advance, conducted with employer knowledge, relying primarily on document review and structured worker interviews on–site – are poorly suited to identifying forced labour. This is not a reason to abandon audits entirely, but it is a reason to understand their limitations clearly and not treat a clean audit result as assurance that forced labour is absent.

Audits can be useful for:

Identifying documentation of wage payments, contracts, and hours that are inconsistent with forced labour indicators
Flagging the presence of recruitment agency arrangements that warrant further investigation
Building a baseline of supplier information against which more targeted investigation can be directed
Identifying physical conditions (housing, worksite access, presence of surveillance or locks) that are visible indicators

Audits are unlikely to surface:

Disclosure of coercion or threats by workers who fear retaliation
Debt bondage arrangements established before workers arrived at the facility
Restriction of movement or document retention not visible on a site walk-through
The experiences of workers who are absent from the site during the visit
Disclosure of coercion or threats by workers who fear retaliation
Debt bondage arrangements established before workers arrived at the facility
Restriction of movement or document retention not visible on a site walk-through
The experiences of workers who are absent from the site during the visit

Worker interviews: the essential but fragile method

Direct conversation with workers is the most important identification method for forced labour – and the one most likely to yield useful information. But it only works when workers have genuine reason to trust the person conducting the interview, believe their responses cannot be traced back to them, and have some confidence that speaking up will not make their situation worse.

This requires:

Independence from the employer. Interviews conducted by auditors introduced by the employer, or in a space controlled by the employer, will not produce honest disclosure from workers at risk. Effective worker interviews are conducted independently, ideally by organisations that have existing relationships with the worker community or credibility in the relevant cultural context.

Confidentiality that is real, not just declared. Workers need to see evidence that their responses are genuinely confidential, not just be told so. This includes conducting interviews away from the worksite, in workers’ own language, without employer representatives present, and with clear explanation of how responses will be used.

Trust built over time. In high–risk contexts, a single visit by an unfamiliar investigator is rarely sufficient to generate disclosure of serious exploitation. Civil society organisations and trade unions that have ongoing relationships in a sourcing community are therefore far better placed to identify forced labour than external auditors making one–off visits.

Grievance mechanisms

Accessible grievance mechanisms – channels through which workers can raise concerns confidentially – are important for ongoing identification of issues that are not visible through periodic audits or visits. But they only function if workers trust them, which requires: channels that are not controlled by the employer, operation in workers’ languages, genuine protection from retaliation, and evidence that concerns raised actually result in action.

For workers in situations of forced labour, grievance mechanisms alone are rarely sufficient – the same power dynamics that prevent disclosure in interviews also prevent workers from proactively raising concerns. They are most useful as a supplement to, rather than substitute for, active investigation.

Collaboration with civil society and trade unions

ETI’s tripartite model – bringing together companies, trade unions, and NGOs – reflects a long–standing insight: the organisations with the best information about working conditions in specific locations and sectors are typically not companies or their auditors, but trade unions with local membership and civil society organisations with established community trust.

Engaging these organisations – as sources of risk intelligence, as partners in worker engagement, and as participants in the design of identification approaches – substantially improves the quality and reliability of forced labour identification. This is not a peripheral activity; in high–risk contexts, it is often the only reliable pathway to disclosure.

Responsible purchasing practices and their role in identification

Identification is not only about looking outward at suppliers. A company’s own commercial practices shape whether forced labour is more or less likely to occur in the first place, and whether it is possible to surface.

Short lead times, last–minute order changes, and extreme price pressure push suppliers to rely on informal labour arrangements and subcontract to facilities that are not subject to any identification efforts. These practices also create conditions in which suppliers have no incentive to be transparent about their labour sourcing.

Building identification into the commercial relationship – making it a normal part of supplier dialogue rather than an external compliance exercise – requires internal alignment between sustainability, procurement, and commercial functions. Companies that approach identification as something that happens to suppliers, rather than something that suppliers and buyers work on together, will find it systematically harder to surface genuine issues.

What to do when indicators are found

Finding indicators of forced labour is not a failure – it is the result of looking properly. What matters is the response.

The immediate priority is the welfare of workers who may be at risk. This means:

Not disclosing your findings to the employer before workers are safe. Tipping off a supplier that forced labour has been identified can result in retaliation against workers who may have spoken to investigators.
Engaging specialist support. Situations involving potential forced labour benefit from involvement of specialists – civil society organisations, trade union representatives, or specialist consultants with experience in worker–centred remediation.
Not defaulting to immediate disengagement. Cutting off a supplier on discovery of forced labour can deprive affected workers of income and access to remedy without addressing the underlying situation. In most cases, staying engaged to support remediation produces better outcomes for workers than exit.

Frequently asked questions

Can a clean audit result mean there is no forced labour in my supply chain? Not reliably. Standard social audits have well–documented limitations in detecting forced labour. A clean audit result is evidence that the audit did not find forced labour using the methods deployed – it is not assurance that forced labour is absent. Effective identification requires methods specifically designed to surface coercion and exploitation, including confidential worker engagement through trusted intermediaries.

How do I know which suppliers to prioritise for forced labour identification? Risk–based prioritisation should focus on the combination of factors that most strongly predict forced labour risk: reliance on migrant workers recruited through agencies, particularly cross–border; sectors with documented forced labour prevalence (agriculture, garments, construction, electronics, domestic work, fishing); countries or regions with weak labour law enforcement or active state–imposed forced labour; and evidence of debt bondage in recruitment practices.

My suppliers don’t want to share information about their labour recruitment practices. What can I do? Supplier reluctance to share recruitment information is itself a risk indicator. Building transparency into supplier relationships requires framing it as a shared responsibility rather than an audit exercise, providing suppliers with support to understand what is expected and why, and using commercial leverage – including purchasing commitments and long–term relationships – to create conditions in which transparency is in the supplier’s interest. Industry–level initiatives and collective action through ETI can also reduce the burden on individual suppliers.

Should we use technology – apps, blockchain, satellite imagery – to identify forced labour? Technology tools can support identification in specific ways: grievance reporting apps can make it easier for workers to raise concerns; supply chain mapping platforms can improve tier visibility; satellite imagery can identify certain agricultural practices. But no technology currently available can substitute for direct worker engagement. The most important information about forced labour – whether workers feel coerced, what recruitment practices were used, whether they are free to leave – can only be gathered through human conversation in conditions of trust.

We identified what looks like forced labour in a Tier 3 supplier we have never directly interacted with. What are our responsibilities? Under the UN Guiding Principles on Business and Human Rights, companies are expected to address human rights risks that are directly linked to their operations, products, or services through their business relationships – regardless of whether they have a direct commercial relationship with the entity where the harm is occurring. Responsibility is graduated by proximity and leverage, but the starting point is to understand the situation and consider what influence, direct or indirect, can be brought to bear to support improvement.

Further reading in this cluster

What is modern slavery? Definitions, forms and scale
Forced labour & modern slavery: a complete guide for business