Two Government Officials Suspended Over AI Policy Plagued By Non-Existent Sources

Two officials within the Department of Communications and Digital Technologies (DCDT) have been placed on precautionary suspension.

The action follows the discovery of fabricated sources in the reference list of the Draft South Africa Artificial Intelligence (AI) Policy. The department had previously published the draft AI policy for public comments, with the closing date of mid-June 2026.

The draft policy proposed the formation of a new AI governance ecosystem. That ecosystem included a National AI Commission, an AI Ethics Board, an AI Regulatory Authority, an AI Ombudsperson, a National AI Safety Institute, and an AI Insurance Superfund designed to compensate individuals harmed by AI systems in cases where liability is unclear.

However, Minister Solly Malatsi was forced to withdraw the policy. It emerged that the integrity of the policy was compromised because it contained various fictitious sources in its reference list.

The Portfolio Committee on Communications and Digital Technologies grilled the minister and the department on the implications of the withdrawal, the way forward, and consequence management on Tuesday.

Officials Suspended And Internal Investigation Launched Over AI Policy Flaws

The minister explained to the committee that once the news expose had emerged, he requested that the director-general, Nonkqubela Jordan-Dyani, launch an internal investigation into the matter.

“As a result of those internal engagements, the secondary step that took action was, she placed the officials on precautionary suspension to allow the team and the department, then she do a thorough investigation around that process, but equally to do it in a manner that both protect the officials, because we have to apply the rules of natural justice, in order to be compliant with the public service regulations and also the country’s labour regulations, as it were, dealing with consequence management,” Malatsi said.

The DCDT said that the matter is still under investigation but is in its final stages. The department said they will update the committee on their progress as the process unfolds.

“Precautionary suspension, I just want to highlight, it does not give the presumption of guilt. It is a process that we need to adhere to, and therefore, we needed to ensure that once the investigation was taking place, we needed to make sure that we could do it without the necessary involvement of the officials as well,” Jordan-Dyani said.

Malatsi said that to remedy the gaps in oversight which led to the AI hallucinations in the policy, they have created a seven-member advisory panel. The panel will help strengthen the integrity of the policy document.

The National AI Expert Review Panel will be chaired by Professor Benjamin Rosman (Wits Machine Intelligence and Neural Discovery Institute) and made up of distinguished experts spanning AI research, law, and governance as follows: Professor Vukosi Marivate, Professor Alison Gildwald, Heather Irvine, Dr Tshepo Feela, Dr Jabu Mtsweni, and Advocate Lufuno Tshikalange.

Chilean Document Translated From Spanish Linked To Fake References

The committee grilled the minister and department on how the error was introduced, the precautionary suspension, and who would be paying for the advisory panel given that the department would already have knowledgeable people housed within.

DCDT Director-General, Nonkqubela Jordan-Dyani said: “So when we reflected on this in terms of where the root cause emanates from… It has been said that ChatGPT was used in as far as the editing of the document itself, but the issue related to the fictitious reference, we’ve also been able to locate and pinpoint where it emanated from.”

It was only later in the meeting that more insight into this arose. Jordan-Dyani said: “There’s an actual document that was shared with the officials, which they used as a context with the drafting of the draft AI policy.

YOU MAY ALSO LIKE: Judge Questions Fannie Nkosi Over R350 000 Cash Found At Home: ‘I Don’t Even Get That Salary’

This document, we’re told, that it’s from Chile and they had translated it as well from Spanish into English, and that is the source from which the information emanates in terms of those references that have been contained in the draft policy,” Jordan-Dyani said.

Committee chair Khusela Sangoni-Diko said, “Just for clarity and as we step off this point, so you’re saying the suspended officials took a document from Chile, which had fictitious references, and then translated it from Spanish to English, and it then found itself in our national AI policy?”

To which Jordan-Dyani said: “As I said, because that’s part of the investigation, but that is the reference point at which we are at.”

Department Admits Weak Internal Checks As Committee Demands Answers

Sangoni-Diko said that she was unsure whether that makes sense. “We have unfortunately run out of time on this matter. I am sure that there are more questions than answers for the members.”

Earlier in the meeting, Malatsi had said that the department fell short in its internal processes for AI authentication and verification. “So that in the event that there aren’t any of those disclosures, because there will be instances where, whether intentionally or unintentionally, people don’t disclose their reliance on AI usage…that they are internal guardrails.”