South Africa Withdraws Draft AI Policy Over Fake References

South Africa's Minister of Communications and Digital Technologies, Solly Malatsi, withdrew the Draft National Artificial Intelligence Policy on Sunday evening, following an internal investigation that confirmed the document contained fictitious academic references, likely generated by artificial intelligence tools without proper verification.

The draft policy, published in the Government Gazette for public comment earlier in April 2026, aimed to establish national priorities and norms for AI development while addressing sector-specific needs. It proposed the creation of several new regulatory institutions, including an AI Ethics Board to oversee ethical standards, an AI Safety Institute to monitor risks, and an AI Insurance Superfund to compensate those harmed by AI-driven outcomes. Public comments were scheduled to close on June 10, but the withdrawal halts the process indefinitely.

The controversy began with a report from News24, which examined the policy's 67 references and found that several cited academic journal articles did not exist or were not published in recognized outlets. Some authors credited with key research had never written on the attributed topics. When...