Artificial Intelligence (AI) is often celebrated for its transformative potential to transform industries, automate processes, and reduce human labour.

However, beneath the shiny cover of progress lies a harsh reality: AI’s efficiency is fuelled by millions of human workers worldwide, toiling for low wages under difficult conditions.

This phenomenon is particularly evident in Kenya, a growing hub for digital labour, where “humans in the loop” carry out vital yet undervalued tasks for major tech companies like Meta, OpenAI, and Google.

A recent 60 Minutes feature, “Training AI Takes Heavy Toll on Kenyans Working for $2 an Hour,” sheds light on the underbelly of AI development, focusing on the Kenyan workforce shouldering this burden.

Kenya’s Role in Training AI Systems

Kenya, often referred to as the “Silicon Savannah,” has become a focal point for tech outsourcing due to its youthful, educated population and low labour costs.

Workers here perform data-labelling tasks essential for training AI systems. This involves tagging images, annotating videos, and even reviewing sensitive content to help AI systems learn to identify objects, faces, or patterns accurately.

For instance, workers on platforms like Remotasks, a digital labour platform operated by Scale AI, often sign up to complete tasks remotely, only to face delayed or withheld payments, citing alleged policy violations.

The promise of AI jobs is appealing in a country grappling with unemployment rates as high as 67% among young people. But the reality is far worse.

READ ALSO:

OpenAI Secures $4 Billion Credit Line

Exploitation or Opportunity? The Dark Side of AI Labour

While tech giants reap billions in profits, the workers enabling this technological leap remain underpaid. In Nairobi, AI labourers hired by outsourcing firms earn as little as $1.50 to $2 (Ksh 195-Ksh 260) an hour, despite reports showing that companies like OpenAI pay intermediary firms such as Sama up to $12.50 (Ksh 1620) per worker hour.

Adding insult to injury, many workers are subjected to short-term contracts, unrealistic deadlines, and high-stress environments. Some even liken these conditions to modern-day slavery.

Narima Wako-Ojiwa, a Kenyan civil rights activist, pointedly calls these workplaces “AI sweatshops,” where computers replace sewing machines, but the exploitation persists.

The Psychological Toll

One of the most horrible aspects of this work is the psychological impact. Workers tasked with filtering graphic content for AI systems describe enduring hours of exposure to disturbing material, including violence, abuse, and pornography. This content moderation leaves lasting scars, including depression, anxiety, and PTSD.

For example, workers hired by Sama to moderate content for OpenAI and Meta recall being tricked into the roles by misleading job descriptions.

One worker shared how the exposure left him unable to maintain healthy relationships or even enjoy aspects of his personal life, such as intimacy with his spouse.

While companies claim to provide mental health support, workers have called the measures inadequate, demanding qualified trauma specialists to address their deep-seated psychological damage.

Legal Battles and Worker Pushback

Kenyan digital workers are beginning to push back against these conditions. Lawsuits against companies like Sama and Meta cite unreasonable working conditions and psychiatric harm.

However, legal protections are limited. Kenya’s labour laws, designed decades ago, fail to address the challenges of digital labour, leaving workers vulnerable to exploitation.

As a result, when workers demand better conditions or governments attempt to enforce stricter regulations, companies often shut down operations and move to neighbouring countries with even weaker oversight.

This creates a vicious cycle, as governments desperate for jobs hesitate to challenge tech giants for fear of losing their investment.

What Needs to Change?

The exploitation of digital workers in Kenya underscores the urgent need for:

1. Fair Wages and Transparency: Companies must pay workers directly and fairly, reflecting the value of their contributions to billion-dollar AI systems.
2. Stronger Labour Laws: Governments must modernise labour regulations to cover digital work and ensure accountability for international companies operating within their borders.
3. Mental Health Support: Companies must provide robust mental health services, including access to licensed trauma specialists.
4. Global Advocacy: International scrutiny is essential to prevent tech companies from exploiting low-income regions under the guise of innovation.

READ ALSO:

Microsoft and OpenAI’s AI Supercomputer: The Rise of Stargate

The Role of Platforms like Remotasks

Digital labour platforms such as Remotasks, where Kenyans frequently find these jobs, play a central role in this ecosystem. While offering a gateway to income, they often leave workers trapped by abruptly shutting accounts or denying payments under unsubstantiated accusations of policy violations.

Despite such platforms’ promises, they exemplify how tech companies sidestep direct accountability, using intermediaries to distance themselves from allegations of exploitation.

Outlook

Kenya’s rise as a global hub for AI labour highlights a glaring inequality in the digital economy. As tech companies boast about their AI breakthroughs, the voices of those powering this progress remain quiet under layers of exploitation and injustice.

True progress lies in ethical innovation where technological advancement uplifts everyone involved, from engineers in Silicon Valley to workers in Nairobi.

It’s time for the tech industry to match its promises of a brighter future with real action toward fair labour practices and equitable treatment.

For more insights into Kenya’s role in AI labor and its human cost, watch “The Hidden Struggles Behind Kenya’s Role in Advancing Artificial Intelligence.” DW Documentary, 14 Nov. 2023, https://youtu.be/qZS50KXjAX0.