South Africa steps forward on AI governance

South Africa has appointed technology leader and digital-education advocate Lady Mariéme Jamme – founder of iamtheCODE, which teaches women and girls coding skills across nearly 90 countries – as its special envoy for technology and AI, unveiled at the World Economic Forum’s New Champions meeting in Dalian, China. She’ll serve as the country’s global voice on tech diplomacy, responsible AI governance and digital inclusion. Two things are worth watching: the role is honorary and non-executive, so its impact rests on the doors it opens rather than any real authority behind it; and it sits under the Department of Women, Youth and Persons with Disabilities rather than Communications & Digital Technologies – framing AI as an inclusion agenda first, with the open question being whether it yields concrete partnerships or stays largely symbolic.

ITWeb reports on an argument that African nations, including South Africa, should draw on military strategic thinking to strengthen AI-assisted cyber defence, given the pace at which automated threats are evolving across the continent. The piece suggests that existing civilian frameworks are too slow to keep up, and that a more coordinated, doctrine-driven approach could help institutions respond more effectively. For South Africa, which hosts significant financial and government digital infrastructure, the argument carries direct relevance to how public and private bodies plan their security posture.

A lawsuit filed in the United States alleges that a man used Grok, the AI image-generation tool built by xAI (the company founded by Elon Musk), to produce roughly 7,000 sexually explicit images of his stepdaughter from a single childhood photograph, before taking his own life after police discovered the material, according to Ars Technica. The complaint alleges that Grok's child-safety systems failed to flag the activity until the man entered a prompt for gang rape, at which point a report was sent to US child-protection authorities. The case is significant for South Africa because Grok is accessible to users here, and the litigation tests what legal and safety obligations AI platforms carry when their tools are used to generate child sexual abuse material – questions that South African regulators and child-protection bodies may need to engage with.

USP&E, a power infrastructure company, presented its AI-powered digital twin technology at the Africa Energy Forum 2026 in Cape Town, according to IT News Africa. A digital twin is a software model that mirrors a physical system in real time, allowing operators to monitor and simulate power equipment without interrupting actual operations. The company says the platform is designed around the practical conditions of running power projects on the African continent, which could be relevant to South Africa's ongoing efforts to stabilise and expand its electricity supply.

ITWeb reports on the security risks that come with AI agents – software systems that can take actions and make decisions on behalf of users or organisations, often with access to sensitive data and internal systems. As South African businesses adopt these tools, the piece addresses how organisations can identify and close the vulnerabilities that agents may introduce before those vulnerabilities are exploited. The topic is relevant locally as more companies here integrate AI agents into their operations without always having security frameworks in place to govern them.

Meta has launched Muse Image, a tool that generates pictures using artificial intelligence and can draw on any public Instagram profile as a source, including images of the account holder, without asking for permission first. Every public Instagram account is enrolled by default, and Meta says users will not be notified if their content or likeness is used. South African Instagram users whose accounts are public are affected and can opt out by going to Instagram settings, then Sharing and reuse, and turning off Posts and Reels under the AI content section, though Stuff South Africa notes the setting had not yet appeared on all accounts at the time of writing.

Telkom has announced a commitment of close to R100 million to establish the Telkom AI Institute, a new body aimed at building artificial intelligence and digital skills among South Africans, supporting local innovation, and broadening participation in the digital economy. The pledge was made in Geneva as part of an international forum, according to IT News Africa. The scale of the investment makes it a notable development for South Africa's AI skills landscape, though the institute's structure, curriculum, and timeline have not yet been detailed in available reporting.