The Department of Home Affairs (DHA) has suspended two officials following the discovery of AI 'hallucinations' within a revised policy paper on citizenship, immigration, and refugee protection. The errors, specifically found in the document's reference list, highlight challenges in integrating AI into official processes.
The DHA acknowledged the embarrassment and will implement AI checks and declarations in its approval workflows. Two independent law firms have been appointed to manage disciplinary processes and review policy documents dating back to November 30, 2022, aligning with the public release of ChatGPT. This incident follows a similar occurrence with the Department of Communications and Digital Technologies, emphasizing the broader need for caution and adaptation in the public sector's use of AI.
