AI and the Administrative State (LAWS90308)

The “administrative state” is the combination of many agencies and bureaucratic institutions exercising the practical power of government. It is traditionally aligned with the executive branch, and is subject to administrative law. But the administrative state is undergoing significant change in many Western liberal legal systems, due to the rapid uptake of algorithmic tools and decision support systems. For example, the United Kingdom’s Home Office used an automated system to make initial decisions on people’s visa applications, in a process that was later found to introduce racist discrimination. Where will these changes lead us? How does the turn to artificial intelligence (AI) further the aims and activities of the executive branch of government? In what ways does AI introduce tensions and conflicts, and how is it connected with broader transformations—intended and unintended? 

In this subject, students will gain knowledge of the technical systems being deployed in the administrative state today, as well as a broad overview of the impacts of algorithmic technologies. These include changes to decision-making, organisational change, the evolution of administrative processes, and involvement in transformative ideas like “government as a platform”. These changes will be considered in light of emerging public law and broader regulatory frameworks, including shifts in public law doctrine and fundamental principles. 

Students will also have the opportunity to consider bigger-picture effects on the rule of law, legitimacy and authority, public values, and changing modes of evaluation that accompany hybrid human-machine systems. The administrative state has always been a hybrid of political, technical and legal form – combining professional bureaucracy, scientific management and efficiency goals with political and normative theory, doctrinal law and resonant social values. Amidst contemporary debates over AI ethics, fairness and safety, the administrative state is a uniquely important site for examining the impact of data-driven algorithmic technologies on democratic institutions and public values.  

We will make regular reference to the Australian content, but examples and issues for discussion will be drawn also from the UK, USA, and the EU. 

Intended learning outcomes

A student who has successfully completed this subject should be able to: 

• Explain the main reasons for the deployment of artificial intelligence and machine learning tools and techniques (AI) in the administrative state;
• Examine how the technical features of AI are deployed within the administrative state, and analyse their interaction with human and other organisational factors;
• Critically appraise technical systems from the point of view of legal sources and materials, including constitutional and legislative frameworks, legal doctrine, and secondary texts including interdisciplinary critical approaches;
• Assess and interpret the main ways that AI's intended and unintended effects challenge the administrative state's normative frameworks, organisational structures and procedures;
• Critically analyse the main ways that the administrative state's organisation, reasoning and evaluation practices are changing in response to the use of AI;
• Hypothesise about how AI might affect the administrative state, including its politico-normative values and justification, in the short-, medium-, and long-term.

Generic skills

• Ability to investigate, evaluate, synthesise and apply legal knowledge in contexts involving the deployment of AI.
• Ability to undertake high-quality interdisciplinary research.
• Ability to make high-quality, well-supported written arguments involving the synthesis of materials from legal and other fields.
• Capacity for self-directed learning, organisation and time management.
• To value and participate in collective discussion, analysis and reflection involving a variety of perspectives.
• Capacity to engage with issues in contemporary society, particularly the social, ethical, and political dimensions of new technologies.