Is technology helping to prevent forced labour, or only document it?

What has worked, and what hasn’t, in the decades-long fight against forced labour? And how can technology, increasingly deployed to root out exploitation, move beyond documenting the problem to helping prevent it? These were among the questions I addressed this week at the annual Engaging Business Forum in New York, drawing from my experience working on forced labour and labour rights over the past 17 years.

In that time, I have witnessed significant changes in how businesses and governments understand and respond to this issue.

We have seen new regulations, from the UK’s Modern Slavery Act, the first national law of its kind to require large businesses to publicly report on forced labour in their supply chains, to the upcoming EU Forced Labour Regulation that bans any product made using forced labour from being placed on, sold in, or exported from the EU market starting 14 December 2027.

We have also seen more robust approaches to human rights due diligence, and an extraordinary expansion of technological solutions. From worker voice apps and satellite imagery to artificial intelligence solutions, companies now have access to tools that would have been difficult to imagine two decades ago.

One of the critical technological developments is the use of data. Used responsibly, better data can help identify risks that might otherwise remain hidden far into supply chains and enable companies and governments to target their interventions more effectively. 

Yet despite this progress, an estimated 27.6 million people were in forced labour in 2021, according to the International Labour Organization. The use of data also has limits. Data-driven approaches can sideline workers' voices, and they tend to focus on documenting abuses, stopping short of prevention or tackling systemic drivers, such as recruitment fees or weak labour protections.

The gap between detection and prevention

One of the most important lessons from nearly two decades of work on this topic is that information is only valuable if we use it to prevent harm, not simply to respond after it has occurred.

Take the example of recruitment fees. Today, companies have better tools to identify the fees migrant workers pay to secure employment, trace recruitment intermediaries and understand the risks associated with particular migration corridors.

When workers have paid recruitment fees, remediation is essential. But if the response stops at reimbursing those workers, we are not necessarily changing the practices that created the problem.

At IHRB, we have long championed the Employer Pays Principle: no worker should pay for a job. The costs of recruitment should be borne by the employer, not the worker.

Putting this principle into practice requires companies to change how they contract and pay recruitment agencies, provide evidence of who has paid for recruitment, examine the incentives driving recruitment practices, and work with governments to strengthen regulation and enforcement. Technology can support all of this, but it cannot make these decisions for companies and governments.

Worker voice requires more than a platform

Digital platforms and apps can make it easier to reach workers, gather information about their experiences and identify concerns that might not emerge through traditional audits or management reporting.

But worker voice is only meaningful if someone is actually listening, understands what they are hearing, and is prepared to act.

Companies and governments often speak about forced labour, trafficking and modern slavery. Workers themselves may not describe their experiences in those terms. They may instead speak about debt, unpaid wages or excessive working hours.

Understanding those experiences requires more than collecting responses via a survey. It requires trust, human interaction and an understanding of the circumstances in which workers live and work. It also requires taking freedom of association, worker-driven monitoring, and independent and effective grievance mechanisms seriously.

We should be asking whether technological interventions are strengthening those relationships or, in some cases, moving us further away from them.

We should also consider what we are asking workers to do. Download an app, complete a survey, identify a violation, provide evidence, label themselves as victims and potentially expose themselves to retaliation?

Workers are not due diligence officers. Their perspectives are essential to understanding risks and designing effective solutions, but companies and governments have ultimate responsibility for identifying and preventing labour exploitation, and to protect and uphold labour rights.

There is also an important distinction between worker voice and worker representation. An app or digital grievance channel may allow individual workers to report concerns, but it cannot substitute for independent trade unions, collective bargaining or trusted worker organisations.

Technology is most promising when it strengthens these existing channels of representation and accountability rather than seeking to replace them.

Forced labour is fundamentally a labour rights issue

There is a tendency to treat forced labour as an exceptional problem requiring increasingly sophisticated methods of detection.

But forced labour does not emerge in isolation. It is closely connected to the power imbalances experienced by workers across global supply and value chains.

Recruitment fees, wage withholding, restrictions on freedom of movement, excessive working hours, discrimination and the absence of effective worker representation can all contribute to situations of coercion and exploitation.

Technology may help us identify patterns and reach workers who would otherwise remain invisible. But preventing exploitation requires strengthening fundamental labour protections: fair recruitment practices, decent working conditions, freedom of association, responsible purchasing practices and effective government enforcement.

Measuring what matters: outcomes for workers

As companies invest in new technologies, we also need to be careful about what we measure and what we reward.

Metrics such as the number of suppliers or recruiters assessed, workers surveyed, risks identified and grievances reported can be useful, but they tell us relatively little about whether conditions are improving.

Returning to recruitment fees, it is important to know how many workers have paid fees and how many have been reimbursed. But it is even more important to know whether new workers are being recruited without having to pay those fees.

Companies should therefore disclose not only what their technological tools have identified, but what has changed as a result. This includes whether those affected can confirm that their conditions have improved. 

Using data to drive collective action

None of this diminishes the potential of technology. On the contrary, better use of data could be transformative, particularly when information is shared and acted upon across companies, sectors and jurisdictions.

This is why IHRB is an insight partner in the Global Data Partnership Against Forced Labour. Launched at the World Economic Forum in 2025, the Partnership brings together businesses, international organisations and civil society to explore how data can be used more effectively to identify forced labour risks and support coordinated action.

By connecting information that is often fragmented across different actors, such collaboration has the potential to reveal systemic risks that individual organisations might otherwise miss.  

Are we designing technology for compliance or for long-term change?

A final concern is that we may increasingly be designing technology to help companies demonstrate compliance, rather than necessarily to change the conditions that make forced labour possible.

Much of the investment in forced labour technology is driven by sometimes competing pressures: regulatory requirements, import restrictions, customer expectations and reputational concerns.

These can create powerful incentives for companies to strengthen their due diligence. But they can also influence what technology is designed to achieve.

A company may invest heavily in demonstrating that a particular product or shipment is not linked to forced labour, without necessarily investing in improving labour conditions across the wider sector or region. The result may be better management of corporate risk, without equivalent improvements in the protection of workers. 

Compliance and prevention should reinforce each other. But we should not assume they always do. 

The question we should be asking is not simply how much more technology can help us see, but what we are prepared to change as a result of what we see.