JOHANNESBURG, April 27 (Reuters) – South Africa has withdrawn its first draft national AI policy after revelations that its bibliography included fictional sources that appeared to be AI-generated.
“The most plausible explanation is that the AI-generated quotes were included without proper verification. This should not have happened,” said Solly Malatsi, Minister of Communications and Digital Technology.
“This failure is not just a technical issue but has compromised the integrity and credibility of the draft policy,” he wrote in a post on Twitter on Sunday.
The policy, released this month for public comment before being finalized, seeks to position South Africa as a continental leader in AI innovation while addressing ethical, social and economic challenges.
It outlines a plan to establish new institutions, including a National AI Commission, AI Ethics Board and AI Regulatory Authority, and create incentives such as tax exemptions, grants and subsidies to encourage private sector collaboration.
Malatsi said there would be consequences for those responsible for drafting the policy, and did not say when the new policy would be released.
He wrote, “This unacceptable omission proves why vigilant human oversight over the use of artificial intelligence is important. It is a lesson we take with humility.”
(Reporting by Nellie Peyton; Editing by Alison Williams)