JOHANNESBURG, May 26 (Reuters) - South Africa's communications minister established an independent panel of experts to review a draft national policy on artificial intelligence, after an earlier version was withdrawn when it was found to contain fictitious and potentially AI-generated references.
Solly Malatsi briefed a parliamentary committee on Tuesday after pulling the earlier draft. The target publication date for public comment is January 2027, according to Jeanette Morwane, the Department of Communications and Digital Technologies acting deputy director-general.
The policy, released in April for public comment, sought to position South Africa as a continental leader in AI innovation while addressing ethical and economic concerns.
Malatsi said internal checks had not flagged the issues before they were exposed by online news publication News24. He added that the draft was meant to be a starting point to invite public input and much of the policy's content had not faced significant challenge. But he acknowledged a "massive oversight" and a lack of disclosure around AI use in compiling the references.
Among other things, the new seven-member panel will review the document, recommend revisions or removals and replace flawed citations, with a revised policy expected to go to the Cabinet by November 2026 for approval, Morwane said.
Two officials have been placed on precautionary suspension pending an investigation. Director-General Nonkqubela Jordan-Dyani said the incident was "highly regrettable," adding that withdrawal was necessary to restore credibility.
(Reporting by Nqobile Dludla; Editing by Thomas Derpinghaus)