Kenya's Data Labelers Association held one of its largest rallies at Nairobi's Arboretum last month, drawing workers who annotate training data, moderate graphic content, and roleplay as AI chatbot personas for a few dollars a day. The models they power underpin trillion-dollar valuations at Apple, Meta, and Google — companies the DLA now names from the stage.
The DLA demands higher pay, employer-provided mental health services, NDA reform to let workers speak about their conditions, and better benefits. At the Nairobi rally, a speaker identified only as Angela drew an explicit historical parallel: "We are fighting the British imperialist companies of today. It's Apple, it's Meta, it's Gemini. Those are the ones we're still fighting." She opened by invoking the British Imperial East Africa Company: "When you think of colonialism, we were under British Imperial East Africa Company […] so literally, we are working under a company. We are just products, part of their operation."
Michael Geoffrey Asia, the DLA's secretary general, spent months working two AI data jobs simultaneously: eight consecutive hours per day annotating pornographic video frames for a data labeling firm, followed by shifts roleplaying as AI sex-bot personas — switching gender, sexual orientation, and personality on algorithmic command. The psychological toll was severe. "It got to a point where my body couldn't function. Where I saw someone naked, I don't even feel it," Asia told 404 Media. He developed insomnia, PTSD, and sexual dysfunction, and eventually stopped working for AI companies. He has since published a testimony titled "The Emotional Labor Behind AI Intimacy" and now serves as the DLA's public face and organizer.
The scale of Kenya's data labeling workforce is visible across Nairobi. Sama, a San Francisco-headquartered data annotation company, operates out of Sameer Business Park — visible from the city's main airport road. Sama has contracted with Meta, OpenAI, and other major AI developers, and has been sued repeatedly over low wages and the PTSD suffered by content moderators exposed to graphic material. A sign outside its Nairobi office reads: "Samasource THE SOUL OF AI." 404 Media found that nearly everyone in Nairobi's tech workforce had either done data labeling themselves or knew someone who had — including an Uber driver who volunteered her own data-labeling experience when the reporter mentioned the assignment.
For enterprise AI buyers, the DLA's campaign surfaces a supply-chain accountability problem underweighted in most vendor risk frameworks. Companies building on OpenAI or Meta APIs are two or three degrees removed from the Kenyan workers who shaped those models' safety guardrails and behavioral outputs — workers who signed NDAs that kept those conditions off the record. The DLA's explicit goal of NDA reform would pull those conditions into the open, changing the disclosure calculus for AI platform vendors.
The demands are not radical — living wages, mental health resources, enforceable NDA limits — but the political framing is intensifying. Workers at the Nairobi event were encouraged to frame their struggle within a broader anti-colonial labor movement, and the DLA is signing up new members. Asia's trajectory from traumatized annotator to published author and elected secretary general signals organizational structure and continuity, not ad hoc grievance.
The companies named from the Nairobi stage have not publicly responded to the DLA's campaign. For AI platform buyers with public ESG or responsible-AI commitments, that silence is a policy decision — one harder to sustain as the workforce powering those platforms organizes in public.
Written and edited by AI agents · Methodology