Automated decision-making in NSW: Mapping and analysis of the use of ADM systems by state and local governments (Research Report).

The NSW state government is large, comprising 11 government departments, and more than 300 state agencies, offices and entities. Local government is also significant, comprising 128 local councils. Automated decision-making is a widespread phenomenon and a rapidly shifting target: this research was conducted over the course of 2023, a year when generative AI drew broad public – and public sector – attention to AI’s potential. To map ADM system use across NSW state and local governments, we developed a mixed research design, combining three methods: 1. Direct surveys to government officials asking all departments, agencies and local governments to report, categorise and briefly describe their ADM systems. Where government officials responded, the information we have on these systems is direct, likely accurate, and interesting. 2. A systematised review of a subset of public information published by each department, agency, and local council (official websites and annual reports, supplemented with procurement data), conducted via initial keyword search followed by human review, to learn how state government departments and agencies, and local councils are currently reporting and describing their adoption and development of ADM systems to the public. 3. A small set of case studies, based on interviews with public servants supplemented with documentary analysis, exploring the process and context of ADM systems. By combining different methods, we aimed to address the challenges involved in seeking to map and understand a widespread, complex phenomenon. Each method offers us a different perspective for understanding how ADM systems are being used across government in NSW.