The problem with letting AI do the grading

As AI increasingly displaces human tasks in education, new research reveals the technology falls dramatically short when it comes to accurately grading student work. A University of Georgia study found that even advanced AI models like Mixtral correctly assess student answers only a third of the time when creating their own rubrics, highlighting the irreplaceable value of human teachers in educational assessment despite growing pressure to automate classroom functions.

The big picture: Teachers are increasingly using AI to grade student assignments as a response to widespread AI use among students, but research suggests this approach fundamentally undermines education quality.

• Nearly 86 percent of university students use some form of AI in their academic work, according to a Digital Education Council study.
• This widespread AI adoption by students has prompted some educators to respond with a “fight fire with fire” mentality, using AI tools to evaluate the same assignments students might be creating with AI assistance.

What some teachers are saying: Various educators have adopted different stances on AI’s role in assessment, ranging from reluctant acceptance to enthusiastic integration.

• One teacher’s philosophy on Reddit crystallizes the reactive approach: “You are welcome to use AI. Just let me know. If you do, the AI will also grade you. You don’t write it, I don’t read it.”
• Other educators are more proactive, using AI to customize learning materials and even requiring students to run their essays through AI systems alongside traditional feedback.

Behind the numbers: University of Georgia researchers tested AI’s grading capabilities and found disturbing accuracy issues when compared to human assessment.

• When tasked with creating its own grading rubric, the Mixtral LLM accurately graded middle school homework only 33.5 percent of the time compared to human graders.
• Even when provided with human-created rubrics, the AI model’s accuracy improved only marginally to just over 50 percent.

Why this matters: The research exposes fundamental limitations in AI’s ability to understand and evaluate student work with the nuance and comprehension that human teachers provide.

• Researchers found that “LLMs can adapt quickly to scoring tasks, [but] they often resort to shortcuts, bypassing deeper logical reasoning expected in human grading.”
• The AI consistently misinterpreted student responses due to its lack of genuine comprehension of the material.

Reading between the lines: The increasing use of AI to grade student work sends a troubling message about how educational institutions value both students and teachers.

• The shift toward automated grading, despite its proven inadequacies, suggests that efficiency is being prioritized over educational quality and meaningful student-teacher interactions.
• With evidence suggesting AI’s comprehension abilities are actually declining, relying on such technology for critical assessment functions risks fundamentally undermining educational outcomes.