South Africa’s first draft of a national artificial intelligence policy has been withdrawn after it was discovered that some citations in the document were fake and appeared to be generated by AI. This mishap is more than just a minor mistake; it highlights the risks of relying on unverified AI output in serious government work. The incident raises questions about how well AI is being managed and scrutinized in policymaking processes.

What Went Wrong with the Draft Policy

The draft policy aimed to position South Africa as a leader in AI development across Africa. It included plans to create new institutions, such as a National AI Commission, an AI Ethics Board, and an AI Regulatory Authority. Additionally, the government proposed offering tax incentives, grants, and subsidies to boost local AI innovation.

However, the draft’s credibility was compromised when a news outlet revealed that some references within the document were fabricated. These false citations not only make research verification difficult but also create a false sense of authority. In contexts like AI ethics and data rights, such falsehoods can seriously undermine public trust and policy effectiveness.

The Dangers of Relying on AI Without Oversight

South African officials suspect that AI tools might have been used to generate these citations without proper human review. Minister Solly Malatsi acknowledged that the integrity of the draft was compromised due to these errors. This incident underscores the importance of human oversight when using AI, especially in critical areas like policy formation.

Using AI without proper checks can lead to embarrassing mistakes and damage a country’s reputation. It’s a reminder that AI should be seen as a tool to assist humans, not replace careful review and verification. Proper supervision acts like a seatbelt—something you don’t realize you need until it’s too late.

What’s Next for South Africa’s AI Efforts

Despite this setback, South Africa is still committed to developing its AI sector. The country has already begun building institutional capacity through its National AI Policy Framework, which was opened for public comment in 2024. This framework aims to explore AI’s economic benefits while addressing governance challenges.

The withdrawal of the draft doesn’t mean the end of South Africa’s AI ambitions. Instead, it serves as a lesson in the importance of careful review and verification. Moving forward, the government will need to ensure that AI-generated content is properly checked to prevent similar mistakes in the future.

Overall, the incident highlights both the promise and pitfalls of AI in policymaking. With proper oversight, AI can be a powerful tool to foster innovation and growth. But without careful management, it risks undermining the very goals it’s meant to support.

Inspired by

• https://ai2people.com/how-ai-policy-in-south-africa-is-ruining-itself/

Sources

• reuters.com
• reuters.com
• news24.com
• gov.za