South Africa’s revised National AI Policy is expected to reach Cabinet by November 2026 and open for public comment in January 2027, according to the Department of Communications and Digital Technologies, setting out a clear timeline for the rebuilding of a process that collapsed earlier this year amid an AI-generated citation scandal.
Jeanette Morwane, acting deputy director-general at the Department of Communications and Digital Technologies, said the updated policy will be sent to Cabinet by November, with public comments invited from January 2027. The schedule provides the first formal post-withdrawal roadmap for the document and its accompanying institutional architecture.
The previous draft, published in April 2026 and approved by Cabinet ahead of public consultation, was withdrawn after reports found fictitious — and in some cases apparently AI-generated — references in the document. The incident damaged the credibility of a policy that was meant to establish how South Africa intends to use and govern AI. “Highly regrettable” was how Director-General Nonkqubela Jordan-Dyani described the withdrawal, saying the move was necessary to protect the department’s credibility.
The revised process is being supported by an independent expert panel appointed by Communications and Digital Technologies Minister Solly Malatsi in May to review the policy, authenticate its references and guide the next phase of drafting. The department has also placed two officials on precautionary suspension while it investigates how the fabricated citations made it into the original Cabinet-approved document.
The withdrawn draft had proposed several new institutions to govern AI in South Africa, including a National AI Commission, an AI Ethics Board and an AI Regulatory Authority, and floated incentives such as tax breaks, grants and subsidies to support local AI development. Those building blocks are likely to remain on the table as the panel and department work toward the November Cabinet submission.
What happens between now and January 2027 will determine more than the credibility of one document. A national AI policy increasingly shapes how AI shows up in hiring, schools, public services, privacy, financial services, healthcare and small business operations. South African companies are already using AI in customer service, recruitment, insurance, banking, media, cybersecurity and software development, while public institutions are exploring AI for service delivery, research, forecasting and administration. Without clear rules, weak oversight could expose citizens to opaque automated decisions, while excessive delays could leave South Africa lagging as other African countries advance their own frameworks.
The fake-citation incident has highlighted a broader truth about generative AI: outputs can look convincing but be entirely fabricated, and human oversight is essential — particularly in government, where policy documents shape laws, budgets, institutions and enforcement for years.
South Africa still wants to be a leader on AI in Africa. The January 2027 draft will be the test of whether that ambition can survive scrutiny — not just on its references this time, but on whether the process behind it can rebuild public trust.
