Opinion | As great powers bet on AI, what of the workforce holding it together?
The future of artificial intelligence depends not only on algorithms but on who controls and benefits from the labour that machines rely on
Artificial intelligence (AI) may look automated, but it runs on human labour. Behind every chatbot and image generator are thousands of people labelling images, tagging text, moderating content and training systems to understand language and culture. This invisible workforce has quietly become a critical layer of the global AI economy.
As AI increasingly enters everyday life, it depends on a vast, largely invisible workforce. By 2022, India’s data annotation industry was estimated to be worth US$250 million, employing 70,000 people, with 60 per cent of its business coming from US companies. Most workers are first-generation annotators; 80 per cent of them are from rural backgrounds.
Most of these centres are staffed by women. In one lab, the entire staff consisted of women, with most of them being college-educated, former stay-at-home wives. Companies like iMerit and Niki.ai have established centres in the states of Odisha, Chhattisgarh and Jharkhand to tap into the Adivasi and other marginalised communities.
This has led to women from marginalised communities being included in the AI industry’s “human-in-the-loop” value chain. Data annotation is often framed as a flexible gig economy job, but it’s also a crucial building block of AI systems. However, many workers have no insight into how their work contributes to the final product.