Can AI adequately assess critical thinking?

New research reveals that AI grading tools can accurately assess certain aspects of students’ critical thinking—such as identifying logical fallacies and evaluating evidence—but they struggle with more nuanced elements like originality and contextual insight. The study, which analyzed thousands of essays across diverse subjects, found that while the technology excels at spotting surface-level errors, it often misjudges deeper analytical depth. Educators caution that relying solely on automated evaluation could overlook subtle strengths in student work, underscoring the need for a blended approach that combines machine efficiency with human judgment.

Though the technology succeeds in evaluating students’ work for some components of critical thought, it falls short on others, research shows.