Who Pays the Price for Our Digital Comfort?

The Invisible Exploitation in Kenya

We can access summaries of hundreds of pages in seconds, learn the content of an hour-long video in a few lines, and accomplish coding tasks that would take a programmer days in a very short time. Artificial intelligence saves time, increases productivity, and sometimes even saves lives through medical diagnoses.

However, behind the "miraculous" level AI has reached today lies an invisible cost. This price is paid with the labor, time, and mental and physical health of people living in less fortunate regions of the world. There is a significant group of people whose mental and physical limits are pushed so that safe content can be produced for users.

The Dark Side of Artificial Intelligence

Almost every day, the short videos we watch on YouTube and the "reels" content we encounter on social media undergo a "data labeling" process to train AI algorithms and filter out harmful content. This process cannot yet be automated. Videos are watched one by one by people who decide what they are about, what they contain, and what they represent. In other words, AI is still learning from humans.

Data labeling is an extremely arduous and grueling task. In addition to the mental and physical fatigue caused by staying in front of a screen for hours and constantly having to make decisions, the psychological exhaustion resulting from exposure to extremely disturbing content is very difficult to compensate for.

In the West, relatively high wages were paid for this job, and the physical and psychological problems caused by data labeling were addressed within insurance systems. However, a new way was found to reduce these costs and risks: to protect the mental and spiritual health of its own society, the West moved this line of work to a geography it has exploited in various forms throughout history: Africa.

Another global power that realized this "solution" was China: China also followed the same path as the West, turning to the African continent, especially Kenya, known as the "technology savannah," which offered more attractive conditions.

Why Kenya?

Despite youth unemployment exceeding 60%, Kenya creates an ideal digital labor pool for global technology companies due to its high education level and English proficiency. The country has the advantage of working in the same time zone as the West and China. Regulations are lax, and worker protection mechanisms are weak. The high unemployment rate forces even qualified young people to accept almost any condition.

Kenya is also marketed as Africa's ideal "technology hub" with its digital service infrastructure and low costs. This time, exploitation is legitimized by the discourse of "digital development" and "youth employment."

The Meta–Sama Case: The First Breaking Point of Visible Exploitation

The picture of digital exploitation targeting Kenya first became visible on a global scale with the Meta–Sama case that erupted in 2022-2023. Sama, a US-based subcontractor handling content moderation work for Meta's Facebook and Instagram, forced hundreds of Kenyan youth it employed in Nairobi to filter extremely psychologically burdensome content for very low wages.

Employees exposed to thousands of images containing murder, child abuse, rape, and extreme violence began to show serious trauma symptoms over time. According to testimonies reflected in court files and the international press, these digital workers were not clearly told about the nature of the content when hired; only vague descriptions like "technology support services" were used. Adequate psychological support was not provided to employees suffering from post-traumatic stress disorder, depression, and anxiety disorders.

Everything was legal on paper. There were contracts, company names were known, offices were official. But thanks to legal loopholes, weak worker protection mechanisms, and high unemployment rates, the resulting picture was a clear system of exploitation in ethical terms.

Jobs that initially started with simple and harmless questions like "how to make coffee?" evolved over time into savage and shocking content like "how to cook human flesh?" or "how to kill a baby?" Digital workers were forced to face the dark side of building AI. Protests were organized, legal avenues were tried; but no concrete gains were achieved.

Digital Exploitation Continues

The form of exploitation changed but did not disappear. Today, Western AI companies post fake ads; job descriptions use phrases like "customer support services." The real job emerges long after the hiring process.

China, on the other hand, follows a much less regulated and legally detached method. A system has been established that operates entirely within closed networks based on referrals, without any legal company structure, contract, or point of contact. In this system, young people labeling videos often do not even receive the wages promised to them. For a job that pays $20 an hour in the USA, Western and Chinese companies offer a Kenyan youth $2.

The process usually starts with WhatsApp groups of about ten people. In these groups, they are asked to watch and label tens of thousands of short videos, each averaging a few seconds. When the weekly quota is not met, groups are closed without any explanation; there is no point of contact to demand rights.

Payments are made via M-Pesa, the mobile payment system widely used in East Africa; but employees are not provided with any payslip, contract, or legal record.

While the West exploits existing legal loopholes, China advances by establishing unregistered digital networks.

Thousands of Videos, Thousands of Micro-Traumas

There is extensive research on the harms of watching short videos. Experts emphasize that consuming short but intense content leads to sudden dopamine spikes in the brain; this weakens long-term concentration ability, dulls abstract thinking, and creates serious cognitive stress. Insomnia, eye and spine problems, chronic tension, and neurological disorders are among the common consequences of this process.

A Kenyan youth forced to watch and label tens of thousands of short videos a day is actually forced to make tens of thousands of micro-decisions. This situation rapidly erodes mental capacity; sleep is postponed to meet the quota, and the body and mind are kept in a constant state of alarm.

Moreover, the content of the videos often includes violence and extreme sexuality. For hours, they watch people having their throats slit, abused children, bodies burned alive, mothers killing their own babies, pornographic content.

Digital workers exposed to violent content for long periods develop serious psychological problems over time.

Digital Awakening

All this picture is transforming Kenyan digital workers not only psychologically and physically, but also politically. They have become not only the exploited but also the first actors to expose this system.

For artificial intelligence to provide more benefits to humanity and become reliable, this grueling work must be done by someone. The problem is that this benefit and reliable platforms are produced through boundless exploitation.

The demands of digital workers are extremely humane: a fair wage, psychological support and guarantees that should be provided after exposure to heavy content, and secure employment contracts.

While companies earn billions of dollars, it is a clear injustice that those doing the most difficult and risky part of the job cannot share in these earnings. Digital exploitation points to a structural problem beyond individual cases.

Similar digital labor networks are rapidly spreading today in countries like Uganda, Rwanda, Ethiopia, and South Africa; Kenya is just a starting point for many companies.

Conclusion

The digital devices we carry in our pockets today are the product of bodies exploited in cobalt and coltan mines in Congo. The "intelligent" systems inside these same devices work thanks to minds and souls exploited in Kenya. In Congo, the ground is excavated; in Kenya, minds are excavated. One is the exploitation of bodies in mines, the other is the trauma of minds behind screens.

The original of this article was published in Independent Turkish.*