What does it mean to graduate “AI-capable” students in political science and public policy? As Artificial Intelligence (AI) becomes embedded in governance, policy design, and public administration, universities are seeking to articulate and teach new forms of capability relevant to this. Yet there remains limited clarity about what “AI capability” entails—particularly beyond technical skills. This paper explores how AI capability is being conceptualised across different public jurisdictions and the implications for politics and policy curricula.
This paper addresses two core questions:
(1) How is AI capability defined across different institutional contexts?
Drawing on comparative analysis of state public sector capability frameworks (including as outlined by the New South Wales and Victorian state governments as well as the Commonwealth), departmental AI policies and responsible use guidelines, and emerging federal approaches, it identifies both convergences and divergences in how capability is framed. While some approaches emphasise technical and data literacy, others prioritise social, ethical, and evaluative dimensions, including judgement, accountability, reflexivity, and the responsible use of automated systems. These differences lead to the second important question about the kinds of knowledge, dispositions, and evaluative practices that should underpin a politics graduate’s formation:
(2) What do these differing conceptions of capability mean for teaching AI-informed decision-making?
If AI capability extends beyond technical proficiency, it requires pedagogical approaches that support students to engage critically with AI as a socio-technical and governance phenomenon. We explore how public sector decision-making frameworks—particularly those embedded in AI policy and regulatory guidance—should or inform curriculum design.
A distinctive feature of this joint paper is its comparative focus on context. It examines how local characteristics—including state jurisdiction, departmental mandates, and institutional priorities—shape approaches to AI capability. It also considers how university-wide commitments, such as those articulated in statements like the Castlereagh Statement, influence program design and teaching approaches. In doing so, it highlights both shared challenges and place-based differences in preparing graduates for AI-enabled governance environments.
The paper will conclude with outlining key capabilities based on a synthesis of state and federal approaches to inform the development of program-level political science and public policy responses that are relevant to local contexts. These will emphasise: the integration of technical, social, and dispositional capabilities; alignment with public sector expectations; and responsiveness to disciplinary and institutional contexts. By bridging insights from policy, practice, and pedagogy, this paper contributes to ongoing debates about curriculum renewal and graduate capability in the age of artificial intelligence.
