(NAROBI, KENYA) – Kenya is developing new labour rules that would require major artificial intelligence companies, including OpenAI and Meta, to meet minimum pay standards and provide mental healthcare for local workers. The proposed policy from the ICT Ministry aims to establish protection guidelines for content moderators and data annotators, who review harmful material and label data to train AI systems like ChatGPT.
The guidelines would mandate that AI firms and their outsourcing partners comply with locally set duty-of-care standards. These include safeguards against exposure to harmful content, access to mental health support, and transparent contracting practices. The government is preparing to publish official occupational protection guidelines that cover minimum standards for written contracts, psychosocial support, grievance procedures, and working conditions.
Kenyan workers employed by outsourcing firms serving global technology companies have for years reported psychological trauma and unfair pay. The draft policy states that the government will support the development of fair and transparent pay standards for the AI and emerging technologies workforce. It also promotes the enforcement of duty-of-care standards, including safeguards against harmful content, mental health access, transparent contracts, proportionate surveillance measures, and accessible grievance mechanisms.
A fair-pay-reference framework would set transparent pay benchmarks for data annotation, content moderation, and AI quality evaluation roles. These benchmarks would be calibrated against international rates for equivalent work. Companies employing Kenyan AI workers would need to disclose their pay structures through a compliance reporting mechanism.
Kenya has become a global centre for AI data annotation and content moderation over the past five years, largely due to its large English-speaking workforce. These roles are critical to developing generative AI systems like ChatGPT, Gemini, and Microsoft’s Copilot, as they rely on human reviewers to train algorithms to recognise harmful content. The algorithms behind these bots depend on vast amounts of human-labelled data.
Technology companies increasingly outsource this work to specialist contractors in countries like Kenya to cut labour costs and create legal distance from employment relationships. By outsourcing, tech firms significantly reduce expenses by paying much lower wages than they would for domestic workforces in the United States or Europe. Markets such as Kenya, India, and the Philippines have high youth unemployment, providing a large pool of workers willing to accept low pay. Using external vendors also allows tech giants to distance themselves from direct responsibility for worker welfare.
Kenyan moderators and annotators working for Meta and OpenAI through outsourcing company Sama have previously raised concerns about severe psychological trauma, low pay, and sudden layoffs. Workers say they are exposed to graphic violence, self-harm, murder, child abuse, rape, necrophilia, bestiality, incest, and non-consensual personal video footage captured by Meta AI smart glasses. They have alleged they received little or inadequate psychological support from the US-based firm. Sama has denied these allegations.
Some content moderators were paid between $1.46 (Sh189) and $3.74 (Sh484) per hour. In the United States, moderators are paid an average of $21 (Sh2,719) to $27 (Sh3,496) per hour. At current exchange rates, $1.46 is approximately £1.13 and €1.34; $3.74 is about £2.90 and €3.43; $21 is roughly £16.27 and €19.28; and $27 is about £20.92 and €24.79.
Kenya’s draft policy notes that data annotation and content moderation workers face unique occupational risks, including exposure to harmful content, insecure working conditions, and limited labour protections. It states that existing frameworks provide insufficient safeguards for mental wellbeing, transparency, and employer accountability. The guidelines would apply to both local and international AI firms operating in Kenya. If implemented, the policy would be one of Africa’s most comprehensive labour protection frameworks specifically targeting AI value-chain workers.