<\/p>\n
South Africa has pulled back its draft national AI policy after officials discovered fabricated academic sources within the document, highlighting the challenges of regulating rapidly evolving technology amid internal verification issues.<\/p>\n<\/div>\n
South Africa has withdrawn its draft national artificial intelligence policy after officials concluded that the document\u2019s references included fabricated academic sources, a lapse that has sharpened scrutiny of how governments use AI while trying to regulate it. Communications and Digital Technologies Minister Solly Malatsi said the mistake was unacceptable and promised that those responsible would face consequences, according to local reporting.<\/p>\n
The draft had only recently been moving through the formal process. According to the South African Government News Agency, Cabinet approved the policy on 25 March 2026 and it was later published in the Government Gazette on 10 April, opening a public comment period that was due to run until 10 June. That timeline means the policy was pulled back before the consultation window had closed.<\/p>\n
News24, as relayed by other outlets, found that at least six of the draft\u2019s 67 academic citations appeared to be AI-generated hallucinations, with some references pointing to non-existent papers in real journals. Chinese state media also reported that the most plausible explanation offered by Malatsi was that AI-generated citations had been inserted without proper verification, undermining the draft\u2019s credibility.<\/p>\n
The policy was intended to extend South Africa\u2019s existing AI framework and set out a broader governance model for the technology. Reporting by Media24 and other publications says it envisaged new institutions, including a National AI Commission, an AI Ethics Board and an AI Regulatory Authority, alongside measures such as tax incentives, grants and subsidies to encourage private-sector participation.<\/p>\n
Its withdrawal leaves Africa\u2019s most industrialised economy without a national AI governance blueprint at a time when policymakers are racing to keep up with fast-moving technology. The episode also raises a more awkward question for the department tasked with regulating artificial intelligence: whether the state has enough internal capacity to police the very systems it wants to oversee.<\/p>\n
Inspired by headline at:<\/strong> [1]<\/a><\/sup><\/p>\n