AI vs. Teacher Feedback: Students Find Both Helpful, But Only One Trustworthy

Students are turning to AI for academic feedback, but who do they trust more? A new study of nearly 7,000 university students found that while AI is considered helpful, teachers are seen as far more trustworthy.

Imagine a student stuck on an assignment. Their lecturer isn't available, or perhaps they're worried about asking a "silly" question. So, they turn to ChatGPT for feedback. In moments, they have an answer and can even ask for more clarification.

This scenario is becoming increasingly common. A new study shows that nearly half of surveyed Australian university students are now using generative artificial intelligence (AI) for feedback on their work.

About the Study

Between August and October 2024, a research team surveyed 6,960 students across four major Australian universities. The researchers wanted to understand how students use AI for learning and, specifically, how they perceive the helpfulness and trustworthiness of feedback from AI compared to their teachers.

The participants came from a wide range of fields, including:

Sciences, Technology, Engineering, and Mathematics (36%) 

Health (24%) 

Humanities and Social Sciences (20%) 

Business and Law (20%) 

The student group was diverse: 57% were women, and 72% were between 18 and 24 years old. The majority were full-time (89%), undergraduate (61%), domestic students (58%) who attended on-campus activities (92%).

Students Find AI Helpful, But Not as Trustworthy

The study found that the use of AI for feedback was split right down the middle, with 49.7% of students reporting they used it to get suggestions, identify strengths and weaknesses, or generate new ideas for their university work.

These students found both AI and teacher feedback to be helpful. In fact, a large majority rated both sources positively, with 83.9% finding AI feedback helpful and 82.2% saying the same for their teachers. However, when compared directly, students rated their teachers' feedback as significantly more helpful overall.

The real difference emerged when they were asked about trust. The study revealed a major gap: 

90.5% of students considered their teacher’s feedback trustworthy, compared to just 60.1% for AI feedback.

AI Provides Volume, Teachers Provide Expertise

A thematic analysis of thousands of open-ended responses suggests that students see AI and teachers as serving different, complementary purposes.

Students reported that, compared to teachers, AI feedback was less reliable, relevant, and contextualized for their specific assignments. The study suggests teachers hold deep knowledge of the course and the learner, allowing them to identify "what matters".

However, students valued AI feedback for its accessibility. They noted its ease of use, speed, and the sheer volume of feedback they could request without feeling like a burden. The information from AI was also seen as more understandable and objective.

The Vulnerability Factor

Seeking feedback can make students feel vulnerable. The study's findings suggest that AI can remove this interpersonal risk. Students described AI feedback as safer and less judgmental, creating a space where they can ask questions they might be too embarrassed to ask a teacher. This aligns with other research showing AI feedback can lower student anxiety.

Many Students Don't Know AI Can Help

While half of the students were using AI for feedback, the other half (50.3%) were not. A key reason, reported by 28.1% of this group, was that they simply didn't know it was possible or didn't know how to do it.

Other reasons for not using AI included a lack of trust in the technology (28.7%) and personal values related to academic integrity or a preference for human interaction (22.5%). The researchers note this is concerning, as all students should be supported in understanding how to access tools that many find useful.

What This Means for Universities

As student participants in the study reported, AI is useful for providing quick, accessible feedback. Teachers, on the other hand, excel at providing the expert, contextualized guidance that fosters deeper understanding. It's a bit like getting medical advice from a qualified doctor versus looking up symptoms online—both might be useful, but you'd trust one far more with something serious.

The study concludes that the future isn't about choosing between AI and humans. Instead, the challenge for universities is to find ways for them to work together. AI can complement educators by presenting helpful, digestible information that is always accessible and free of personal judgment. This allows teachers to lean into their strengths: providing the challenging, relational, and expert advice that truly helps students learn and grow.

This AI-co-created article reports on the research paper "Comparing Generative AI and teacher feedback: student perceptions of usefulness and trustworthiness" by Michael Henderson, Margaret Bearman, Jennifer Chung, Tim Fawns, Simon Buckingham Shum, Kelly E. Matthews & Jimena de Mello Heredia. https://theconversation.com/uni-students-are-using-ai-to-ask-stupid-questions-and-get-feedback-on-their-work-263535