Compiled by the editorial desk with reference to the Wired investigation and public statements from the companies and workers cited.

The artificial intelligence tools that power everything from search engines to chatbots rely on vast amounts of human-labeled data, but the people who perform that labeling often work for pennies, according to a report by Wired. In countries like Venezuela, the Philippines, and even refugee camps in Lebanon and Kenya, workers hired through gig platforms such as Appen, Clickworker, and Scale AI frequently earn less than the local minimum wage, sometimes taking home just a dollar or two per day.

One worker, Oskarina Fuentes, a Venezuelan national who moved to Colombia to escape poverty, told Wired that she works 18-hour days from her bed for Appen, an Australian data services company. On a good month, she nets about $280, which is just shy of Colombia's monthly minimum wage of $285. Fuentes described waking up as early as 2 a.m. to get first pick of tasks, which pay between 2 and 50 cents each. In a typical hour and a half, she earns about one dollar, though when work is slow, she may make only a dollar or two per day.

Another worker, an 18-year-old from Pakistan who goes by Mutmain, started working on Appen at age 15 using a family member's ID. He now works from 8 a.m. to 6 p.m., then again from 2 a.m. to 6 a.m., yet often earns no more than $50 per month. “I need to stick to these platforms at all times,” he told Wired, “so that I don’t lose work.”

Because these micro-gig platforms pay only for completed tasks, workers say that research and preparation time often goes uncompensated. Mutmain explained that a task that takes five or six hours of real time might pay only $2, which he described as “digital slavery.”

An Appen spokesperson told Wired that the company is taking steps to reduce the difficulty workers face in finding tasks, but must strike a “careful balance” between the rapid completion of tasks that clients want and the consistency of work that laborers need.

Why This Matters

The report highlights a hidden cost of the AI industry: the human labor behind algorithm training is often outsourced to regions with high unemployment and low wages, where workers have little bargaining power. While companies like Microsoft, Google, and Amazon rely on this labeled data to improve their AI systems, the workers who make it possible are often paid far below a living wage.

The situation echoes broader concerns about gig economy labor practices, where platforms classify workers as independent contractors, avoiding minimum wage and benefit obligations. In the United States, similar debates have surrounded prison labor, where inmates earn pennies per hour for work that often benefits private companies.

As AI continues to expand into everyday products, the report raises questions about the ethical responsibilities of tech companies to ensure fair compensation throughout their supply chains. Appen and its competitors have not yet announced any major changes to their pay structures, but the Wired investigation adds to a growing body of evidence that the AI boom may be built on the backs of some of the world's most vulnerable workers.