Unseen Labour: The Human Cost of Artificial Intelligence
AI runs on hidden human labour. Ghost workers face low pay, harsh conditions, and exploitation—demanding urgent ethical reforms.
This blog is based on today’s editorial “Unseen Labour, Exploitation: the Hidden Human Cost of Artificial Intelligence” by Nivedita S., The Hindu, 17 September 2025 and is useful for UPSC Mains answer writing, especially GS Paper II, III, and GS Paper IV (Case Studies). Artificial Intelligence (AI) is often seen as fully automated, but behind every system are invisible “ghost workers.” These workers, mostly in developing nations, face low wages, disturbing tasks, and poor conditions while training AI. Their exploitation raises urgent ethical and policy concerns.
Notes-Making
Main Idea: AI appears automated, but it actually relies on hidden human workers—mostly in developing countries—who face exploitation, poor pay, and harsh working conditions.
Key Points:
- Human Labour and AI
- AI cannot process raw data on its own.
- Data annotators/labellers mark images, audio, video, and text to train AI models.
- Example: Teaching AI what the colour “yellow” is, or helping a self-driving car recognise people.
- Better AI depends on better data, which requires more human work.
- Training Process of LLMs
- Large Language Models (LLMs) like ChatGPT are trained in three steps.
- Humans are key in the second and third steps: fine-tuning, giving feedback, and correcting errors.
- Working Conditions
- Big tech firms outsource to countries like Kenya, India, and the Philippines.
- Workers earn very low wages (less than $2/hour) and work long hours.
- Many must view disturbing content (violence, pornography), which can cause PTSD and other mental health problems.
- Lack of Expertise and Transparency
- Some jobs need expert knowledge (e.g., labelling medical scans), but companies often use non-experts, causing errors.
- Workers usually do not know which company they serve.
- Work comes through online gig platforms, making responsibility unclear.
- Workers are paid per microtask, constantly monitored, and easily fired.
- Resistance and Suppression
- Kenyan workers protested, calling their jobs “modern-day slavery” and accusing US firms of breaking labour laws.
- Those who complained or tried to unionise were dismissed.
- Conclusion & Call to Action
- These hidden workers are called “ghost workers” and are vital to AI.
- The author urges stricter laws to ensure transparency, fair pay, and dignity in the AI supply chain.
Summary
The article argues that the “automated” nature of Artificial Intelligence (AI) is a myth, as it is fundamentally built upon vast amounts of hidden human labour. This work, known as data annotation, involves people labelling raw data—such as images, text, and video—to train AI systems like ChatGPT and self-driving cars. The better the AI, the more human effort is required to create its training data.
However, this crucial work is largely outsourced by large Silicon Valley tech firms to workers in developing nations like Kenya and India. These workers face severe exploitation, including poverty-level wages (often under $2 an hour), exposure to traumatic content that damages their mental health, and a complete lack of job security. Furthermore, the labour supply chain is deliberately fragmented through online gig platforms, making it opaque and hard to regulate. The author concludes that the advancement of AI is powered by this modern-day exploitation and calls for urgent legal reforms to protect these invisible “ghost workers” and ensure ethical practices in the AI industry.
Response
Nivedita S.’s article shines a light on an issue often ignored in discussions about artificial intelligence (AI). While AI is usually praised for being fast, efficient, and seemingly independent, the article reminds us that behind every system are real people doing hard work. By showing this hidden side, the author pushes the debate beyond technology to questions of fairness and justice.
One of the article’s strongest points is how it explains AI in simple terms. Instead of describing AI as a machine that learns on its own, the author shows how it depends on people. Workers have to label data, check mistakes, and even decide what is harmful content. For example, large language models like ChatGPT cannot even recognise a colour like yellow unless people have labelled data to train it. Social media platforms also need human moderators to filter graphic or harmful content. These examples make complex processes easy to understand for a general audience.
The article is also powerful because it focuses on the workers themselves. Many of them are in countries such as Kenya, India, and the Philippines. They often face long hours, very low pay, and little recognition. Some workers even described having to review violent and disturbing images, which can cause serious mental health problems. By including such testimonies, the article turns abstract discussions about “AI ethics” into real stories of exploitation and struggle.
Another strength is how the article links this problem to global inequalities. It points out that while tech companies in places like Silicon Valley make huge profits, much of the labour is outsourced to poorer nations. This shows how technological progress is not neutral but shaped by power, wealth, and geography. The author challenges the idea that AI is simply a universal good, arguing instead that it is built on unequal systems that need to be exposed.
The article also makes an important call for change. It argues that ethical AI is not only about protecting data privacy or reducing algorithmic bias but also about improving the working conditions of those who support these systems. The writer calls for stronger laws, fair wages, mental health support, and the right to unionise. This focus on workers’ rights is an important addition to debates about AI governance.
One striking phrase used in the article is “ghost workers.” This highlights how these labourers are invisible both to the public and to the companies who benefit from their work while avoiding responsibility for their welfare. In some cases, their treatment has been compared to “modern-day slavery,” which is a serious claim. It raises the question of whether we can call progress “progress” if it relies on pushing vulnerable people into degrading work.
The biggest lesson from this article is that true innovation should not only be measured by how fast or smart machines are. It should also be judged by fairness and humanity. AI should not advance at the cost of hidden suffering. Workers’ rights, fair pay, and dignity must be central to discussions about the future of technology. By drawing attention to these issues, Nivedita S.’s article serves as both a warning and a call to action. It reminds us that the story of AI is not just about machines but also about the people who make them work.
Case Study Question (Ethics Paper)
Artificial Intelligence (AI) is often presented as fully automated, but in reality, it relies on thousands of human workers—commonly called “ghost workers”—mostly in developing countries. These workers perform data annotation, content moderation, and training tasks for big tech companies.
Reports suggest that they are often underpaid (sometimes less than $2 per hour), exposed to disturbing and harmful content (such as violence and pornography), and denied basic rights like fair wages, safe working conditions, and the right to unionise. When workers have protested, companies have dismissed them or dismantled their unions.
The author of a recent article argues that ethical AI should not be limited to ensuring data privacy or reducing algorithmic bias, but must also include the protection of workers’ dignity, rights, and wellbeing. The article calls for stricter laws, fair pay, mental health support, and international regulation of AI labour supply chains.
Questions:
- What are the ethical issues involved in the working conditions of these AI labourers?
- Identify the stakeholders and their respective responsibilities in this context.
- As a policymaker in a developing country hosting such workers, what moral dilemmas would you face in balancing foreign investment, economic growth, and the protection of workers’ rights?
- Analyse this case using the principles of:
(a) Utilitarianism
(b) Deontological ethics
(c) Virtue ethics
- Suggest a comprehensive way forward to ensure ethical AI development while safeguarding the dignity and rights of workers.
Answer:
The case presents ethical concerns around the hidden human workforce, often called “ghost workers”, who sustain Artificial Intelligence systems.
Ethical issues: The workers face exploitation through poor wages, job insecurity, and harsh conditions. Their dignity is compromised when they are forced to view disturbing content without adequate mental health support. Lack of recognition and denial of collective bargaining rights further violate principles of fairness and justice.
Stakeholders:
- Workers – deserve fair pay, safe conditions, and dignity.
- Tech companies – have a duty to ensure humane labour practices and transparent supply chains.
- Governments of host countries – must enforce labour laws and protect rights.
- Consumers – indirectly benefit and must demand ethically developed AI.
- International institutions – should frame global digital labour standards.
Moral dilemmas for policymakers: Balancing economic growth and foreign investment with workers’ rights; creating employment opportunities while preventing exploitation; and ensuring competitiveness without compromising ethics.
Application of ethical theories:
- Utilitarianism: Exploitation cannot be justified as the suffering of many outweighs the benefits of efficient AI.
- Deontology: Workers must be treated as ends in themselves, not merely as means.
- Virtue ethics: Compassion, fairness, and integrity demand respect for human dignity.
Way forward: Governments should mandate minimum wages, enforce labour rights, and provide mental health support. Tech firms must disclose their labour practices and respect the right to unionise. International collaboration is needed for global AI labour standards.
Conclusion:
True progress lies not in faster machines but in ensuring that technological innovation respects fairness, justice, and human dignity.
Subscribe to our Youtube Channel for more Valuable Content – TheStudyias
Download the App to Subscribe to our Courses – Thestudyias
The Source’s Authority and Ownership of the Article is Claimed By THE STUDY IAS BY MANIKANT SINGH
