Algorithmic decisions in education governance: implications and challenges

Abstract

Abstract In this perspective article, I explore the implications of artificial intelligence (AI)-enabled algorithmic decisions on education governance. Three main questions are explored: (1) Are algorithmic decisions de facto policy decisions? (2) What distinct features of algorithmic decisions necessitate a re-evaluation of education governance? (3) How should one begin addressing algorithmic decisions in education governance? The analysis suggests, first, algorithmic decisions can indeed be considered de facto policy decisions, as they are often made by private companies with substantial public consequences but little oversight. Second, three distinct features of algorithmic decisions—fast speed of development and implementation, lack of interpretability and transparency, and unpredictable emergence of new capabilities—call for a re-evaluation of education governance. Third, to address algorithmic decisions in education governance, I propose a proactive approach to multilevel social control mechanisms, which includes federal and state legislation, local enforcement, non-governmental organizations, and individual stakeholders. The discussion in this perspective article will stimulate conversations that scrutinize how AI, and algorithmic decisions specifically, challenge traditional assumptions of education governance, including the separation of powers, power distribution between national and local governments, due process, and representative democracy. The discussion aims to shed light on the evolving landscape of education governance in the age of AI.

Keywords

• Artificial intelligence
• Education governance
• Education policy
• Decision making