Sama to lay off 1,108 Nairobi staff after Meta contract exit

Sama, Kenya's largest AI data annotation and training company, announced redundancy notices affecting 1,108 Nairobi-based employees following the termination of a major contract with Meta Platforms. The decision, formalized through Section 40 of Kenya's Employment Act 2007, marks a significant contraction for a firm that has positioned itself as a critical infrastructure player in Africa's emerging artificial intelligence ecosystem.

The company's statement reveals that management engaged in negotiations to preserve the workstream but ultimately failed to secure a continuation arrangement. This outcome underscores a fundamental fragility in Kenya's tech employment landscape: heavy dependence on a small number of multinational clients operating on contract-based models with limited commitment to long-term regional capacity.

**Context: The Meta-Sama Relationship**

Sama has been a cornerstone contractor for Meta's content moderation and AI training operations across Sub-Saharan Africa. The partnership exemplified how Western tech giants leverage African labor pools to handle computationally-adjacent tasks—data labeling, content review, and training dataset preparation—that require human judgment but command significantly lower wage bills than equivalent European or North American operations. For Kenya, Sama represented one of the country's most visible success stories in the global digital economy, employing thousands of educated, English-speaking workers.

The contract termination suggests Meta is either consolidating operations, automating previously manual workflows, or redistributing work to alternative vendors in lower-cost jurisdictions. Without explicit confirmation, the most probable scenario involves workflow consolidation as Meta's in-house AI capabilities mature, reducing the need for external annotation contractors.

**Market Implications for European Investors**

This development carries three critical lessons for European entrepreneurs and investors evaluating African tech infrastructure plays:

**First, client concentration risk is acute.** Sama's over-reliance on a single megacap client created existential vulnerability. For investors evaluating African SaaS, BPO, and AI services companies, due diligence must scrutinize customer concentration. A single client representing >40% of revenue should trigger red flags regardless of growth metrics.

**Second, automation poses a permanent structural headwind.** Meta's contraction reflects a broader trend: as machine learning models improve, the cost of high-quality training data falls, and fewer human annotators are required. European investors betting on labor arbitrage alone—the traditional value proposition of African tech services—should expect margin compression as automation deepens. Competitive advantage must shift toward domain expertise, specialized knowledge, and intellectual property rather than pure cost savings.

**Third, Kenya's regulatory framework functioned as intended.** The Employment Act 2007 enforcement here actually demonstrates institutional maturity, not weakness. Sama followed proper redundancy procedures, suggesting Kenya offers genuine legal protections. However, this also means labor cost savings don't remain permanent—severance obligations and regulatory compliance consume margins on contract terminations.

**Forward Outlook**

Sama faces a critical pivot. The company can either: (1) diversify its client base aggressively, targeting enterprise customers beyond Big Tech; (2) integrate vertically into AI product development, moving beyond labor services; or (3) shrink to sustainable scale. The 1,108 redundancies represent roughly 30-40% of Sama's estimated Nairobi workforce, suggesting the company is choosing controlled contraction rather than restructuring.

For European investors, this signals that African AI services remain viable—but only as a component of a broader, diversified strategy. Pure-play labor arbitrage plays are increasingly vulnerable to automation and client concentration.

---

#