Grading in the Age of AI: Equity, Transparency, and the Role of Human Feedback

As artificial intelligence continues to shape higher education, automated grading software has become a more widely adopted—but often unquestioned—tool. While marketed as time-saving and objective, these platforms raise critical questions about equity, transparency, and pedagogical values. This presentation will draw on Neil Selwyn’s call for a “critical but balanced” lens in evaluating classroom AI tools and will examine how automated grading reshapes the fundamental practices of teaching and learning in higher education while also paying close attention to equity-related issues. Our session will share findings from a research project piloted in an English composition classroom. We will evaluate two commonly used grading platforms, comparing them to human feedback across dimensions such as linguistic diversity, writing style, and rhetorical complexity. Through a combination of literature review, a classroom trial, and student/instructor feedback, we explore how these tools may reinforce algorithmic bias or disadvantage multilingual and non-standard English writers. At the same time, we hope to identify moments where automation may support learning when used thoughtfully alongside instructor input.