Jobs, data, and who gets to decide

MTN Group chief executive Ralph Mupita has been appointed to a new United Nations commission on artificial intelligence, according to TechCentral. He joins a panel that includes Rwandan President Paul Kagame, Salesforce founder Marc Benioff, and Nvidia chief Jensen Huang. The appointment places a prominent South African business leader at the table where global AI governance norms are being shaped, giving the continent a direct voice in those discussions.

TechCentral publishes an opinion piece arguing that South Africa's ability to benefit from AI will depend on deliberate policy choices rather than chance, describing the opportunity as requiring what it calls 'statecraft'. The piece is relevant to ongoing South African debates about how government and institutions should respond to the spread of AI, though the editor may wish to assess whether the framing sits comfortably with the newsletter's position that South Africans should choose where this technology takes them, rather than feel pressed to keep pace with it.

Writing in Business Day, commentator Heath Muchena argues that the debate over whether AI company valuations form a bubble is missing a key factor: the significant expansion of the US money supply, which inflates asset prices across the board and makes headline figures harder to interpret at face value. The piece suggests that before concluding AI is overvalued or fairly priced, investors and institutions need to account for what the underlying currency is actually worth. For South African investors and analysts watching AI market signals, the argument is a caution against reading dollar-denominated valuations without adjusting for monetary context.

A survey of 1,864 young adults aged 18 to 35 across ten countries in Africa and South Asia, conducted by global charity Restless Development and researchers from Makerere University and the University of Cambridge, finds that young people in lower-income countries are broadly more optimistic about AI than their peers in the United Kingdom, with nearly 80% expecting it to improve their education and learning. Half of respondents expected AI to improve their employment prospects within five years, yet significant barriers remain: in most African countries surveyed, more than half cited high data costs as a constraint, and many flagged that widely used AI tools do not support their languages, with respondents from Zimbabwe and Uganda noting that tools built around major world languages marginalise languages such as Shona and Luganda. The findings carry direct relevance for South Africa's debates on digital inclusion and AI governance, particularly the survey's finding that respondents overwhelmingly prefer governments, rather than technology companies, to set the rules on AI regulation and privacy.

Writing in Business Day, an opinion contributor argues that AI risks deepening South Africa's existing inequality, pointing to the country's lag in technology investment as a key factor. The piece raises concern that without deliberate intervention, the benefits of AI will concentrate among those already better resourced, while those with less access fall further behind. The argument is opinion rather than new research, but the question it poses – who in South Africa stands to gain, and who does not – is one the publication will return to as evidence accumulates.

IT News Africa reports that AI is reshaping how careers start rather than simply eliminating jobs outright, and argues that South Africa's experience is likely to differ from global patterns given the country's distinct labour market conditions. The piece pushes back on both the replacement narrative and the straightforward productivity story, suggesting the real effects will be more uneven and context-specific. For South Africa, where youth unemployment is already high, how AI changes entry-level hiring and early career pathways is a question with significant consequences for workers and employers alike.

South African tech publication Hypertext reports that a wave of mass job cuts across the global technology industry in 2026 is being driven, at least in part, by companies adopting AI to automate work previously done by people. Microsoft's latest round of redundancies is among the most prominent examples; the company's chief people officer acknowledged that AI is changing how work gets done, even while stating that the roles cut are not being directly replaced by the technology. According to employment-tracking site Layoffs.fyi, nearly 120,000 jobs have been cut across roughly 219 organisations so far in 2026, a pace that is running ahead of 2025's full-year total and that has implications for how South African labour policy and workers respond to AI adoption in their own workplaces.

Google has quietly changed how it handles user data across Search, Maps, Translate, and other services, with users now opted in by default to having their activity, including images and audio, used to train Google's AI systems. This is a shift from the company's previous approach of relying on publicly available information. South African users of these services can limit this by turning off the new Search Services History and Personalised Recommendations settings in their Google accounts, as Hypertext (htxt.co.za) explains step by step.