The content that keeps platforms and AI systems usable - filtering out graphic violence, abuse, and explicit material - is reviewed by human moderators, and a long-running Kenyan case put their working conditions on the legal record. Facebook content moderators based in Nairobi, employed through the outsourcing firm Samasource Kenya EPZ (Sama), alleged poor working conditions, inadequate mental-health support, low pay, and unfair dismissal after raising concerns. Meta argued that as a foreign company it could not be sued in Kenya’s Employment and Labour Relations Court (ELRC).
On September 20, 2024, the Court of Appeal of Kenya, in Meta Platforms, Inc & 2 others v Motaung & 186 others ([2024] KECA 1262), dismissed Meta’s appeals and let the case proceed in the Kenyan courts. The court treated Meta’s jurisdiction challenge as turning on contested facts rather than a clean point of law, holding that “whether or not the appellants are engaged in virtual business is an issue of fact” that “could not be determined by way of an interlocutory application.” The practical effect was that 186 moderators could pursue their claims against both Meta and Sama at trial.
The ruling is significant because it is one of the first times a court allowed the people who perform AI- and platform-adjacent content moderation to litigate directly against the technology company whose product they cleaned, despite the layer of outsourcing in between.
Why business readers should care: content moderation and data labeling are routinely outsourced to contractors in lower-cost jurisdictions, and companies often assume that contracting structure insulates them from liability. This case shows that assumption can fail - the conditions of the human workforce behind an AI or platform product are an enforceable legal and reputational exposure, not just a vendor’s problem.