South Africa’s AI Policy Withdrawn Over AI-Generated Fake Research: A Wake-Up Call for Digital Governance. JOHANNESBURG, SOUTH AFRICA – The South African Department of Communications and Digital Technologies has withdrawn its Draft National AI Policy after uncovering that it contained AI-generated fake research citations.
The policy was gazetted on April 10th but was pulled just 16 days later following revelations of non-existent journals and fabricated authors in its bibliography. This incident has underscored the necessity for rigorous human oversight in AI-related policy-making and the importance of addressing the unique context of Africa’s digital landscapes. Officials commented on the matter.
The draft policy aimed to establish oversight bodies such as a National AI Commission and an AI Insurance Superfund to compensate citizens harmed by AI — driven outcomes.
However, the revelation of AI — generated fake research highlights the potential pitfalls of over-reliance on AI in critical decision-making processes. Dr. Nomalanga Mashinini, a senior lecturer in cyber law, argues that this incident should not be framed as an embarrassment but as an opportunity to examine the broader implications of AI in governance.
She emphasizes the need for accountability, transparency, and explainability in AI governance, which are non — negotiable conditions for responsible AI development.
The withdrawn draft policy aimed to position South Africa as a leader in AI while addressing data sovereignty, ethical AI deployment, and local innovation capacity.
However, the scandal has raised questions about the effectiveness and reliability of AI tools in shaping national policy.
As the South African government moves forward with a revised draft, it will be crucial to address the lessons learned from this incident. Ensuring that human oversight is a central component of AI policy-making and promoting transparency and local expertise will be key to building public trust and fostering responsible AI development in the country. This incident serves as a cautionary tale for other countries as they consider incorporating AI into their policy-making processes.
The need for human oversight and the inclusion of local expertise in AI governance will become even more critical as the world increasingly turns to AI to inform and shape its policies.
*Additional reporting by ImNews | Sources consulted: 5*
—
This original article was produced by the ImNews editorial team
Source: Nomalanga Mashinini
