AI hallucinations in two government documents prompt suspensions and calls for verification policy

South Africa's Draft National AI Policy had to be fully rewritten after AI-generated fabricated sources were discovered. A separate Cabinet-approved immigration white paper faces the same problem, with two officials suspended.

Categorized in: AI News Government
Published on: May 04, 2026
AI hallucinations in two government documents prompt suspensions and calls for verification policy

South Africa's Government AI Failures Expose Cabinet Oversight Gaps

South Africa's Draft National AI Policy contained fabricated sources generated by an AI tool, forcing a complete redraft of the document meant to guide the country's AI use.

Minister of Communications and Digital Technologies Solly Malatsi revealed the hallucinations last week. The ANC Study Group on Communications and Digital Technologies called it "one of the most alarming failures of ministerial oversight and intellectual rigour in the recent history of South Africa's digital governance."

The problem extends beyond the policy draft. The Department of Home Affairs suspended two senior officials after discovering that AI had fabricated sources in the Cabinet-approved Revised White Paper on Citizenship, Immigration and Refugee Protection.

The department said the hallucinated references were appended to the document after the fact and never cited in the main text. The reference list has been withdrawn pending an independent investigation.

Cabinet Approval Process Raises Questions

The real issue isn't the references themselves-it's that Cabinet members approved a document without verifying them. High-ranking officials tasked with checking documents before publication failed to catch the fabrications.

Leon Schreiber, DA Coordinator in the National Executive, announced that Democratic Alliance ministers will now require AI verification as part of policy document approval. He plans to raise the issue at the next Cabinet meeting to implement this across government.

Verification Tools Have Their Own Problems

AI detection tools often produce false positives, flagging legitimate work as AI-generated. The effectiveness of these tools for cross-referencing local academic journals remains unclear.

If AI was used in these two documented cases, it's likely being used elsewhere in government. Officials will continue using AI to draft important documents until the national policy is finalized and implemented.

For government employees tasked with policy development and document review, understanding AI for Government applications and limitations is now essential. Learning proper prompt engineering practices can help prevent similar failures in your department.


Get Daily AI News

Your membership also unlocks:

700+ AI Courses
700+ Certifications
Personalized AI Learning Plan
6500+ AI Tools (no Ads)
Daily AI News by job industry (no Ads)