Last month, South Africa’s Communications minister announced that he had decided to withdraw the country’s national AI policy after discovering that at least 6 of its 67 academic citations turned out to be AI hallucinations, mentioning journal articles that didn’t exist.

“The most plausible explanation is that AI-generated citations were included without proper verification. ​This should not have happened,” Solly Malatsi said. “This ⁠failure is not a mere technical issue but ​has compromised the integrity and credibility of the ​draft policy.” 

The National Artificial Intelligence Policy sought to position South Africa as a leader in AI innovation, while attempting to address ethical, social, and ‌economic ⁠challenges, but only managed to embarrass the South African government after media outlets discovered that a significant part of the document had been created by AI.

“This unacceptable lapse proves why vigilant ⁠human ​oversight over the use of ​artificial intelligence is critical. It’s a lesson we take with humility,” the South African Communications Minister added.

South Africa’s News24 was the first to report that the African nation’s AI policy, which had just been unveiled for public comments, included obvious AI hallucinations, including cited articles that didn’t really exist from journals such as the South African Journal of Philosophy, AI & Society, and the Journal of Ethics and Social Philosophy.

Mr. Malatsi said that there would be “consequences” for those responsible for drafting the ​national AI policy, and added that the document will be revised before it is reissued for public comment.