Digital Shadows: Unveiling the crisis of forced labour in the tech age

Mail & Guardian

12-06-2025

(Graphic: John McCann/M&G)
The digital age is characterised by a fragile balance between promise and peril. The peril is often imagined through the lens of dystopian fiction — tales of mass surveillance,
Forced labour is prohibited by the International Labour Organisation (ILO) Forced Labour Convention 29 of 1930. According to the ILO, forced labour refers to any work performed under coercion, threat or without free and informed consent.
At a national level, South Africa's Constitution provides a clear prohibition — section 13 states that 'no one may be subjected to slavery, servitude or forced labour'. This is supported by domestic legislation such as section 48 of the Basic Conditions of Employment Act, which gives further effect to this right.
Despite these protections, enforcement mechanisms have not kept pace with the complexities of the developments introduced by digital platforms and transnational digital labour markets.
According to an ILO report,
Digital forced labour does not resemble the traditional imagery of chained factory workers. It is more insidious, woven into the underpaid and often invisible work of data labellers, content moderators, AI trainers and even cryptocurrency miners.
These people, often based in low-income regions, are employed under precarious arrangements with little oversight. They might work 12-hour shifts for a few dollars a day, with no formal contract or labour protection. In South Africa, as in other parts of the continent, this form of labour is particularly concerning given the high rate of youth unemployment, reliance on informal work and rapid growth of gig and platform-based jobs.
The global nature of technology platforms compounds the problem. Many of these companies operate across borders, placing them beyond the jurisdiction of local labour laws and regulatory oversight. A 2019 Forbes article by Adi Gaskell described this vividly: 'It's helped to create a world in which the haves are increasingly well off, while the have-nots make do with insecure and poorly paid work.'
For African workers doing digital piecework, there is often no path to upward mobility — only the repetition of tasks for marginal pay under opaque systems.
Despite the promise of digital innovation, human labour remains fundamental. Many digital platforms still rely on people to label training data, moderate harmful content, provide feedback and enrich machine learning systems with contextual knowledge. Human involvement remains essential not because machines are incapable, but because people are cheaper, more adaptable and more vulnerable. They can be paid little, monitored extensively and replaced easily — a dynamic that meets the ILO's criteria for forced labour, especially when individuals are driven by poverty and lack of alternatives.
One emerging technology that demands closer scrutiny is blockchain. Often praised for transparency, traceability and decentralisation, blockchain also poses a hidden risk when misused.
In certain digital labour platforms, blockchain-based contracts — or 'smart contracts' — are used to automate payment and task allocation. While this can reduce overhead costs, it can also entrench exploitation. Because smart contracts are irreversible and automatically enforced, a worker can be locked into a task with no ability to renegotiate terms, challenge unfair compensation or withdraw.
In some cases, blockchain-based platforms are designed to obscure employer identities, making accountability nearly impossible. Even worse, digital wallets tied to such platforms can require workers to complete tasks before unlocking payment, no matter how unreasonable the demands. This can create a new form of coercion — one that is enforced not by human bosses but by immutable code. In unregulated environments, this form of 'algorithmic servitude' is indistinguishable from digital forced labour.
So what is to be done?
Regulatory responses must evolve. The ILO has recommended a comprehensive approach centred on prevention, protection, remedies, enforcement and planning. Strategies include awareness campaigns, victim support services, stronger legal frameworks and improved data collection.
Critically, the organisation found that ending forced labour would not only raise wages and protect workers but would also reduce unfair competitive advantages for unethical businesses. The economic and social case for action is therefore strong.
But digital forced labour presents new legal challenges. A 2024 article by Thomson Reuters examined this in the US context. It found that although trade laws such as the Tariff Act effectively block goods produced by forced labour, they are poorly equipped to address intangible inputs like datasets or AI-generated models trained through exploitative means. The article argues that new legal instruments are required to extend labour protections to the digital sphere.
Africa faces its own specific issues. In a 2020 analysis, Tshilidzi Marwala noted that although countries like Kenya provide the labour backbone for many global digital operations, they receive little of the value or technological infrastructure. 'While Africa plays a significant role by way of providing cheap labour,' he wrote, 'the continent reaps little from the industry.'
This is a warning that also applies to South Africa. Without intentional investment in regulatory frameworks and infrastructure, the country risks being a perpetual digital labour reserve — supplying skills but not sharing in the dividends of innovation.
What is needed now is a continental vision that safeguards labour rights while advancing Africa's technological sovereignty. South Africa's progressive legal foundation offers a strong starting point. But more must be done to update national frameworks for the platform economy. This includes introducing legal definitions of digital work, mandatory due diligence for companies sourcing digital labour and binding transparency requirements for AI developers.
In parallel, public policy must invest in digital upskilling, ethical AI education and mechanisms to track the provenance of data used in machine learning.
Moreover, regional cooperation is essential. The African Union's Continental Strategy for AI is a promising step but it requires a sharper focus on labour rights, ethical sourcing and enforcement. Institutions such as the Southern African Development Community can also play a role in harmonising policies across borders, creating a collective front against exploitative digital practices.
As consumers of digital products, we seldom reflect on the hidden labour behind our tools — the countless workers whose keystrokes and decisions support our seamless experiences online. But therein lies the real danger — a system that exploits precisely because it remains invisible.
South Africa, and the continent more broadly, must not only expose this system but lead the way in dismantling it — replacing shadows with accountability and peril with promise.
Letlhokwa Mpedi is the vice-chancellor and principal of the University of Johannesburg and a labour law scholar. Tshilidzi Marwala is the rector of the United Nations University, UN under-secretary general, and a thought leader on AI and governance.
The authors' latest book on this subject is Artificial Intelligence and the Law (Palgrave Macmillan, 2024).