Damilare Dosunmu and Tessie Waithira report in Rest of World on Kenyan workers labeling China’s AI videos

Damilare Dosunmu and Tessie Waithira report in Rest of World on Kenyan workers labeling China’s AI videos

On 4 December 2025, Damilare Dosunmu and Tessie Waithira report in Rest of World that Chinese AI companies are hiring young Kenyans to label videos through informal networks of middlemen and WhatsApp groups. Workers interviewed by the publication described short-term projects, no formal contracts, demanding output targets, and uncertainty about which companies use their work. Some said they could lose income for failing accuracy requirements; one worker described becoming “a zombie.” The reporting presents a sharp contrast: data labeling offers income to young people facing unemployment, while the opaque arrangements can leave workers without clear employers, protections, or recourse over pay. Data Labelers Association president Joan Kinyua calls the model “the height of digital colonialism”; Qhala founder Shikoh Gitau says the work could provide earnings but requires policies to protect workers. Rest of World reports that Kenyan authorities are developing regulations, while accountability for workers between outsourcing firms and contracting companies remains under discussion. Payal Arora argues that cheap annotation subsidizes AI development and warns that without fair labor practices, its future will be “fundamentally unjust.” The reporting leaves open how the proposed rules will identify accountable employers and protect workers.

Publication: 4f1824d9-6679-1a9b-26a4-3d657227ab9f

attribution — pass: The reading correctly credits Damilare Dosunmu and Tessie Waithira of Rest of World. Quotes are attributed to the right speakers: Kinyua of the Data Labelers Association, Gitau of Qhala, Arora of Utrecht University and Kimata. This is a single reported article, not a listing.

By Damilare Dosunmu and Tessie Waithira

summary — pass: The summary makes clear who is hiring whom, how the work is organized (middlemen, WhatsApp groups, no contracts) and the stakes for workers: income set against opacity and a lack of protections. It keeps the sharp voices of Kinyua and Arora, including the worker's 'zombie' line. 'Data labeling' is not defined, but its meaning is clear from context. The question follows directly from Kimata's account of unresolved accountability.

“You get into the zone and zone out. You become a zombie,” he said.

date — pass: The stated date of 4 December 2025 matches the dateline and the originalPublished value. Nothing in the reading uses hindsight. The government's regulation plans are described as ongoing work, not outcomes. The page was modified in July 2026, but every passage the reading relies on is plainly part of the original article body. This is a publisher page, so no archive disclosure is required.

4 December 2025 • Nairobi, Kenya

support — pass: Each claim is supported by the cited text: informal hiring, no contracts, two-week projects, accuracy-linked pay, Kinyua's 'digital colonialism', Arora's 'silent subsidy', and regulations still being developed with accountability unresolved. The point about recourse over pay is supported by Kinyua. The cited passages are core article content, so the later page edit does not cast doubt on them.

At times, these workers don’t get paid because they don’t have a direct link with the organization they’ve been working for or what their work is helping to build.

media — pass: No media was selected, and no verified media candidates were supplied, so leaving media empty is appropriate.