South Africa withdraws draft national AI policy after fictitious references are confirmed
The Facts
- Communications and Digital Technologies Minister Solly Malatsi withdrew South Africa’s draft National Artificial Intelligence Policy after concerns were raised about its references.
- An internal review or investigation confirmed that the draft policy’s reference list contained fictitious sources.
- Malatsi said the fictitious citations were most plausibly AI-generated or included without proper verification.
- The draft policy had been released in April 2026 for public comment before it was withdrawn.
- Malatsi said the problem was not just technical and had compromised the integrity and credibility of the draft policy.
- The draft policy was intended to set South Africa’s AI priorities and provide a framework for addressing AI-related ethical, social, economic, or regulatory issues.
- Officials have said an accountability or consequences process is underway to determine responsibility for the flawed draft.
Context
What was wrong with the draft AI policy?
The document’s reference list included fictitious academic sources, and the minister said the most plausible explanation was that AI-generated citations were included without proper verification Yahoo News,IOL,SABC News - Breakin….
Why does the withdrawal matter?
The draft was meant to help define South Africa’s national approach to AI, including priorities, safeguards, and governance questions, so its withdrawal pauses that policy process and raises questions about how the final framework will be rebuilt Yahoo News,Citizen,SABC News - Breakin….
What happens next?
The public-comment process on this draft has effectively been halted by the withdrawal, and Malatsi has said internal accountability steps are underway to determine responsibility and consequences before the policy process moves forward IOL,SABC News - Breakin…,SAnews.
How left and right read it
What stands out is a breakdown in public stewardship around a policy meant to set national AI priorities and address ethical, social, economic, and regulatory issues. If fictitious, plausibly AI-generated citations made it into a draft released for public comment, that raises concerns about whether the public can trust the institutions meant to build safeguards around a technology with broad social consequences.
The key issue is procedural credibility and accountability. Withdrawing a draft after an internal review confirmed fictitious sources was the right step, because a framework for AI policy needs to be grounded in verified material, and the stated process to determine responsibility matters if government is going to exercise authority in this area.
The receipts — all 20 sources
Wire services (3)
Independent coverage (17)
Facts first. Then every angle.
The day’s biggest stories in one short brief — the facts everyone agrees on, then the competing values behind the headlines. Free in your inbox.