A new segment of digital labour is expanding as individuals license their voices, images, and other personal attributes to technology firms in exchange for small, usage-based payments. Compensation can be minimal, with some platforms offering base rates as low as $0.02 per minute for voice data.

This model reflects a broader shift toward monetising personal data as a resource for developing and refining digital systems, while raising questions about long-term value distribution and worker protections.Bouke Klein Teeselink, an economics professor at King’s College London, characterised this trend as part of a growing “gig AI training” economy, where individuals perform fragmented, task-based work tied to data generation.

He noted that companies are increasingly choosing to compensate contributors directly rather than relying exclusively on publicly scraped material, in part to reduce the risk of copyright disputes. This shift also aligns with the need for more controlled and higher-quality datasets.

Veniamin Veselovsky, a researcher in the field, said that human-generated data remains critical for improving system outputs, particularly in areas where existing datasets fall short. He stated that “human data, for now, is the gold standard” for extending system capabilities beyond established patterns.The growth of these marketplaces is closely linked to global economic disparities.

Workers in developing countries, where unemployment is high and local currencies are often volatile, are among the most active participants. Payments in U.S. dollars can provide relatively greater purchasing power, making even low-paying digital tasks financially attractive compared to local alternatives.

For many individuals, this work represents a pragmatic response to limited employment opportunities rather than a long-term career choice.Participants often include individuals who have struggled to secure stable employment or entry-level positions in traditional sectors. In some cases, the income generated is used to fund education or vocational training.

A data trainer based in Cape Town, identified as Louw, said the earnings, while inconsistent, enabled him to save for a $500 course to train as a masseur. He reported difficulty accessing formal employment due to a long-term nervous disorder and viewed the platform-based work as a necessary interim solution. Louw acknowledged the trade-offs involved but emphasised that earning in U.S. currency provided a meaningful financial advantage in his local context.

In higher-income countries, participation is also increasing, though driven by different pressures. Rising living costs have led some individuals to monetise personal data as a supplementary income source. In such cases, the decision is often framed as a financial adjustment rather than a primary occupation, reflecting broader changes in labour markets and household economics.

Despite the apparent accessibility of this work, the contractual frameworks governing these platforms have drawn scrutiny. Many marketplaces require contributors to grant irrevocable, royalty-free licenses over their data, allowing companies to use, modify, and commercialise the material indefinitely without further payment.

This creates a disconnect between the one-time compensation received by workers and the potentially long-term commercial value derived from their data.For example, a brief voice recording could be incorporated into automated systems that operate for years, generating revenue without additional compensation to the original contributor. Similar concerns apply to image and video data, where likenesses may be repurposed across multiple contexts.

The absence of ongoing royalties or profit-sharing mechanisms has raised questions about fairness and sustainability within the model.Transparency is another significant concern.

Participants often have limited visibility into how their data will be used or where it may appear. This lack of clarity increases the risk of unintended applications, including use in contexts that contributors may find objectionable.

Legal protections are also limited, particularly in cross-border scenarios where jurisdictional challenges can complicate enforcement.Mark Graham, a professor of internet geography at the University of Oxford and author of Feeding the Machine, said that while the income generated can be meaningful in the short term, the broader structure of the work presents systemic risks.

He described the sector as “precarious, non-progressive and effectively a dead end,” noting that it does not typically provide pathways for skill development or career advancement. Graham also pointed to what he termed a “race to the bottom in wages,” driven by global competition among workers and the absence of standardised pay structures.

He added that demand for such data may be temporary, shaped by current technological requirements rather than long-term labour needs. As systems evolve, reliance on human-generated inputs could decline, leaving workers without stable income streams or transferable skills.

In this scenario, the enduring value is captured primarily by platform operators and firms based in higher-income economies, while contributors receive only short-term compensation.Personal accounts from participants highlight both the opportunities and the limitations of this emerging form of work.

Coy, who previously licensed his likeness for use in promotional content related to medical supplements for pregnant and postpartum women, described mixed feelings about the experience. He said the process felt impersonal, with public reactions focusing on his physical appearance rather than his identity.

Coy indicated that his initial decision was influenced by a perception that such data would be collected regardless, making compensation preferable to uncompensated use. However, he later expressed discomfort with the lack of control over how his image was used and interpreted.

He has since chosen not to participate in similar opportunities and stated that he would only reconsider if offered significantly higher compensation and clearer terms.His experience reflects a broader reassessment among some participants, particularly as awareness grows around licensing conditions and downstream uses of personal data.

While the market continues to expand, these concerns suggest that its long-term trajectory may depend on evolving standards around transparency, compensation, and worker protections.

A 23-year-old rider, Mohammad Mohsin, said his daily earnings have fallen from around 1,500 rupees ($5.38) to about 1,100 rupees as petrol prices climbed above 320 rupees per litre. “Before it was all mine. Now, a huge sum goes into petrol,” he said.

The spike in fuel prices is hitting low-paid urban workers particularly hard, with many riders struggling to cover basic expenses, let alone seasonal spending associated with Eid. Reuters spoke to more than half a dozen riders across Karachi, all of whom said higher petrol costs had significantly reduced their take-home earnings.

The squeeze comes despite increased demand for deliveries during Ramadan evenings and the Eid period, when food orders typically surge.

Pakistan’s inflation rate, which peaked at 38% in 2023, has eased but remains elevated at around 7%, continuing to push up the cost of food, rent and utilities. Many gig workers earn below the monthly minimum wage of 32,000 rupees ($114.67), with no fixed salaries or employment benefits.

Riders said the combined impact of inflation and fuel costs has eroded their purchasing power, forcing them to prioritise essentials over customary holiday spending.

Companies reliant on gig workers, including foodpanda, say they are adjusting pricing and taking steps to support rider earnings.

However, workers say these measures have not kept pace with rising costs, leaving them to compensate by spending more hours on the road.

Austerity measures, school closures and increased work-from-home practices have also reduced ride-hailing demand, cutting into incomes even as delivery volumes rise.

“We cannot afford clothes or shoes,” said Hizbullah, a 26-year-old delivery rider, reflecting a broader sentiment among workers that a typically festive period has become financially strained.