
In a significant setback for South Africa's digital governance ambitions, the Department of Communications and Digital Technologies has officially retracted its first draft national policy on Artificial Intelligence. The decision follows the revelation that the document contained numerous fictitious citations and references, which were evidently generated by an AI tool without proper human verification. Communications Minister Solly Malatsi announced the withdrawal, emphasizing that the inclusion of fabricated data has severely compromised the credibility of the policy and the department's integrity.
The draft policy was originally intended to serve as a strategic roadmap to position South Africa as a regional leader in AI innovation. It proposed a comprehensive framework to tackle complex ethical and social challenges while fostering economic growth through technological advancement. Among the key initiatives outlined were the establishment of a National AI Commission and an AI Ethics Board to oversee the deployment of these technologies. Additionally, the policy suggested creating various incentives to encourage collaboration between the government and the private sector to accelerate the country's digital transition.
However, the discovery of 'hallucinated' sources has sparked a national conversation about the risks of automated content generation in public administration. Minister Malatsi highlighted the irony of a policy designed to regulate AI being undermined by the misuse of the very technology it sought to govern. The Minister has underscored the necessity for rigorous human oversight in the use of AI tools and indicated that there will be consequences for the individuals responsible for the drafting process, as the department seeks to restore public trust.
This incident serves as a stark reminder of the limitations of Large Language Models (LLMs) and the critical need for fact-checking in high-stakes environments. While the South African government remains committed to developing a robust AI framework, the current draft will now undergo a thorough revision process to ensure all data and citations are accurate. The move underscores a growing global consensus that while AI can assist in administrative tasks, human accountability remains indispensable in the creation of national policy and legislative frameworks.
This story touches markets covered on Anansi Intelligence ↗.
Continue exploring similar stories