Generative AI and the social functions of educational assessment

Abstract

This paper examines the transformative potential of Generative AI (GenAI), particularly large language models such as ChatGPT, for the social functions of educational assessment. Using Patricia Broadfoot’s ‘four Cs’ framework–competence, competition, content, and control–the analysis considers how GenAI may enhance or challenge these traditional roles. Two additional ‘C’s are introduced: ‘credibility’, concerning the authenticity of learner performance in an era of AI-generated outputs, and ‘connoisseurship’, relating to the ability to effectively use and critically assess AI-generated content. While GenAI offers significant opportunities for personalised learning and more nuanced assessments of competence, it also presents challenges related to authenticity, surveillance, and inequality. We conclude that the ‘four Cs’ framework remains a valid lens for understanding the social functions of educational assessment, now augmented by the dimensions of credibility and connoisseurship.