The Australian Bureau of Statistics' most recent version of the internal Artificial Intelligence (AI) Use Case Register or equivalent document that records in-scope use cases

Written by Claude Haiku 5.5 from the extracted text on 20 Oct 2026. It can miss or misread redacted and scanned pages. Check the source before relying on it.

This release is an extract from the ABS internal AI Use Case Register, a corrupted and largely unreadable table of entries with a legend of risk levels (Low, Medium, High), technology types (Generative AI, Machine Learning, Natural Language Processing, Computer Vision) and lifecycle stages (Discovery, Delivery, Live, Closed). Entries are dominated by garbled or placeholder text, so most project descriptions cannot be reliably interpreted. Identifiable items include a Microsoft 365 generative AI tool for staff productivity, a machine learning model for coding Census data (described as in development, with the note that it is not yet run in a secure environment), an employment imputation model using decision trees, and a natural language tool for analysing survey feedback. Risk ratings are mostly Low, and several entries record that models are not yet generated or that human review is required. Privacy and data security safeguards are only sparsely noted, and the register contains no clear figures beyond unexplained numbers such as 10,000,000. The document is of limited analytical value without the source register.