South Africa has pulled back its draft national artificial intelligence policy after the discovery of fabricated academic references raised questions about the document’s integrity.
The policy, released earlier this month for public comment, was intended to outline the country’s strategy for regulating and developing artificial intelligence technologies. It included proposals to establish new governance structures such as a National AI Commission, an AI Ethics Board, and a dedicated regulatory authority.
In addition, the draft aimed to stimulate private sector involvement through incentives including tax relief, grants, and subsidies — reflecting a broader push to position South Africa within the global AI economy.
Policy credibility undermined by AI-generated references
Concerns emerged when analysts identified that several references cited in the document did not exist. Out of 67 academic sources listed, at least six were found to be fabricated, likely generated by artificial intelligence systems.
Communications and Digital Technologies Minister Solly Malatsi acknowledged the issue, stating that the most plausible explanation was that AI-generated citations had been included without adequate verification.
He stressed that the problem extended beyond a minor administrative error, describing it as a failure that compromised the credibility of the entire policy framework.
In media interviews, Malatsi was candid about the situation, calling it both serious and embarrassing. He also pointed to the irony that the policy itself sought to establish ethical standards for the responsible use of AI.
Warnings over over-reliance on AI tools
The incident has reignited debate around the risks associated with the growing use of AI tools in professional and governmental settings.
Malatsi highlighted the absence of sufficient human oversight as a key factor behind the error. He warned that excessive reliance on automated systems, without proper checks and balances, could lead to significant consequences.
The minister said the episode should serve as a lesson for policymakers and institutions adopting AI technologies, reinforcing the need for robust verification processes.
South Africa, like many countries, has been working to develop frameworks to manage the rapid expansion of AI while balancing innovation with ethical considerations. Similar policy efforts have been seen globally, with governments attempting to address issues such as accountability, transparency, and data governance.
The withdrawal of the draft policy is expected to delay South Africa’s formal approach to AI regulation, although officials have indicated that a revised version will be developed.
Authorities have also signalled that there may be consequences for those involved in drafting the document, though no timeline has been provided for the release of an updated policy.
Source: Department of Communications and Digital Technologies
