South Africa Withdraws AI Policy Draft over Fake References

In less than one month after it unveiled its draft national policy on Artificial Intelligence, South Africa has withdrawn the document following revelations that it contained fictitious sources in its reference list, widely believed to have been generated by AI tools without adequate human verification.

Announcing the withdrawal on social media, South Africa’s, minister of Communications and Digital Technologies, Solly Malatsi, said “the most plausible explanation is that AI-generated citations were included without proper verification.

This should not have happened.”
The policy, which had been opened for public comment, was intended to serve as a cornerstone for positioning Africa’s most industrialised economy as a continental leader in AI development, while also setting guardrails around ethical use, data governance and economic inclusion.

However, the credibility breach has cast a shadow over that ambition, highlighting the very risks the policy itself sought to regulate.

“This failure is not a mere technical issue but has compromised the integrity and credibility of the draft policy,” Malatsi added in a post on X, underscoring the gravity of the lapse.
Before its withdrawal, the draft policy had proposed an expansive governance architecture, including the establishment of a National AI Commission, an AI Ethics Board and an AI Regulatory Authority. It also outlined incentives such as tax breaks, grants and subsidies aimed at accelerating private-sector participation and innovation.
While these proposals were seen as progressive, the controversy has now shifted attention from policy ambition to process integrity.
Malatsi disclosed that there would be consequences for officials involved in drafting the document, signalling possible disciplinary action, though he did not provide a timeline for the release of a revised version.
“This unacceptable lapse proves why vigilant human oversight over the use of artificial intelligence is critical. It’s a lesson we take with humility,” he said.
The development serves as a cautionary tale for other African countries, including Nigeria, which are currently exploring national AI frameworks. It reinforces the need for rigorous validation systems as governments increasingly integrate AI tools into policymaking processes.