It is an uncomfortable reality, but modern slavery is endemic in global supply chains. Estimates put the number of modern slavery victims at almost 50 million globally, including those who are subjected to conditions of forced labor. Of these individuals, 17.3 million are being forced to work by private businesses (and nearly 4 million in forced labor imposed by state authorities). And more than 3.3 million of those in forced labor are children.
The extent of the problem – coupled with the complexity of global supply chains – means that businesses (and consumers) in all sectors face a real risk of supporting modern slavery, knowingly or not. Yet, the proportion of large firms detecting and reporting modern slavery activity in their supply chains is low. For example, in its recent analysis of a sample of large firms with operations in the United Kingdom, investment company CCLA found that only 30 companies admitted to detecting modern slavery practices (or indications that it is present) in their supply chains.
Recent analysis by the BBC of tomato purees sold by UK supermarkets illustrates the scale of the problem. While several UK supermarkets advertise their own-brand tomato purees as “Italian,” the report concluded that most of these products are likely to contain tomatoes grown in China. Much like cotton, tomatoes have been linked to the forced labor of Uyghur and other largely Muslim minorities whom the Chinese state views as a security risk.
> It is worth noting: The UK’s food-labelling regulations allow products to be labeled as originating from a particular country, such as Italy, as long as some aspect of processing occurs there.
Faced with the evidence, denial is the natural response from companies due to reputational risks and the outrage that media coverage attracts. Unsurprisingly, the supermarkets named in the BBC investigation have disputed the findings. Herein lies the problem with reporting and transparency as a tool for eradicating the modern slavery cycle: The view that corporate disclosure offers a solution to the modern slavery plague rests on the adage that “sunlight is the best disinfectant.” Transparency promotes scrutiny, which in turn drives out bad practice.
Accordingly, many jurisdictions require companies to report on modern slavery risks in their business and supply chains. In the UK, for example, section 54 of the Modern Slavery Act requires companies with UK operations and a turnover of at least £36 million to publish an annual statement on modern slavery and human trafficking. In practice, however, the average quality of reporting falls short. In particular, while companies do a good job of describing general policies and processes, detailed analysis of clear targets for detecting evidence of slavery is scarce. Instead, the majority of companies adopt an “all’s fine – nothing to see here” approach to reporting.
Inconsistency between the scale of modern slavery activity and low detection rates raises serious doubts over reporting as an effective tool. One possibility is that management is not looking hard enough. Another is “risk-washing” – where companies try to conceal or play down the risks modern slavery poses to their supply chain and business success. Either way, the effectiveness of the self-reporting governance model breaks down.
This is not surprising given the reputational risks facing companies and shareholders. Investors, consumers, and the media, alike, routinely greet evidence of modern slavery activity in a supply chain as unambiguous bad news. In contrast, no news implies such practices are entirely absent.
Really nothing to see here?
Given these reporting incentives and outcomes, it makes sense that self-detection remains the exception rather than the rule. And while inferring no risk from no detection is reasonable when dealing with rare events, it is not appropriate in the face of an endemic problem like modern slavery.
Against this background, the reporting focus and responsibility needs to be flipped. Disclosures should attract appropriate scrutiny, but not outrage and shame by default. Detection is evidence of a well-functioning due-diligence process. Conversely, “nothing to see here” disclosures should attract the skepticism and scrutiny they deserve. This is particularly true when the reporting organization’s suppliers operate in high-risk sectors (for example, agriculture or construction) and high-risk regions. Companies claiming a clean bill of health should be able to show evidence of rigorous due-diligence processes.
Companies also need practical tools to improve their detection and reporting practices. To support this shift in expectations, new technologies offer promising solutions.
Our research using large language models demonstrates how artificial intelligence can help transform modern slavery reporting from a box-ticking exercise into a meaningful assessment tool. These models can analyze company statements with accuracy that resembles human expert evaluation. A combination of better incentives and better tools could finally make supply chain reporting work as intended.
The path forward requires a fundamental shift in how we judge corporate behavior. Modern slavery is endemic, and detecting it should be the norm, not the exception. When a supermarket finds forced labor in its tomato supply chain – oran apparel retailer uncovers labor exploitation in the factories of its suppliers, this should be seen as evidence of robust due diligence, not corporate failure.
Companies operating in high-risk sectors – including those in fashion – must be required to demonstrate concrete detection efforts or provide detailed explanations for finding nothing. Most importantly, investors, media and consumers need to reward rather than punish transparency. Without this shift in incentives, we perpetuate a system where willful blindness is the rational choice for businesses – and millions of victims remain hidden in plain sight.
Mahmoud Gad is an Associate Professor of Accounting and Finance at Lancaster University.
Steven Young is a Professor of Accounting at Lancaster University.
Share
