Do AI feedback tools work better alone or with a teacher?
The clearest finding across these studies is that combining AI feedback with human teacher feedback produces the best outcomes. A randomized controlled trial of 88 intermediate English learners compared three feedback approaches: teacher e-feedback, AI-generated feedback (using ChatGPT and Grammarly), and a hybrid model. The hybrid group showed the most substantial improvements in writing, especially in task achievement, coherence, and grammatical accuracy [3]. Similarly, a study of 86 university students in Tibet found that combined teacher and AI feedback significantly improved writing skills compared to either alone, and the authors noted AI's potential to promote educational equity [1]. These results suggest that AI alone is not a replacement for teachers, but a powerful supplement that can amplify human feedback.
Can AI feedback stand on its own without widening gaps?
Yes, but only under specific conditions. A study of 60 undergraduates using an immersive virtual reality environment for presentation practice found that AI-only feedback was just as effective as teacher-supported AI feedback for improving presentation skills and reducing anxiety [5]. Students in the AI-only group highly valued the feedback and perceived it as an independent, useful source. However, this worked because the AI was well-designed and students were motivated. In contrast, a study of 32 preservice physics teachers using an early (non-generative) AI feedback system found that while feedback acceptance was high, the AI feedback was perceived as less personalized and less useful than human feedback [2]. The key difference is that modern generative AI tools (like ChatGPT) can provide more nuanced, context-aware feedback than older models, but they still lack the personalization and depth that teachers bring. If AI tools are deployed in under-resourced schools without teacher support or adequate digital literacy training, students who struggle may fall further behind.
Does AI feedback boost engagement for all students?
AI feedback can increase engagement, but the effect may not be uniform. A study of 48 Spanish EFL students found that those who received AI feedback enhanced with augmented reality (AR) showed significantly higher behavioral, emotional, and cognitive engagement, as well as increased intrinsic motivation and autonomy [4]. Students valued the immediacy, visual immersion, and personalization of the AR-AI environment. However, this study used a relatively small sample and a novel technology combination (AR + AI), which may not be accessible to all schools. The risk of inequality is that students with less access to technology or lower digital skills may not experience these engagement benefits. The Tibet study [1] explicitly raised AI's potential to promote educational equity, but this potential is only realized if AI tools are designed to be inclusive and are paired with teacher guidance. Without deliberate effort, AI feedback could widen the gap between well-resourced and under-resourced classrooms.
About These Sources
This answer is built on 5 peer-reviewed studies — published from 2024 to 2026, 5 from 2024 or later, 2 in Q1 journals — selected as the most relevant from 5 studies that passed quality screening, drawn from 53 papers retrieved from a database of over 500 million.
Sources used in this answer
Interaction and dialogue: Integration and application of artificial intelligence in blended mode writing feedback
In a quasi-experiment with 86 university students in Tibet, combined teacher and AI feedback significantly improved writing skills compared to either alone, highlighting AI's potential to promote educational equity.
When AI Feedback Was Still in Its Infancy: An Exploratory Comparison of Early AI Feedback Attempts on Preservice Physics Teachers’ Reflective Writing
In an exploratory study of 32 preservice physics teachers, an early non-generative AI feedback system showed no systematic improvement in reflective writing quality, and AI feedback was perceived as less personalized and less useful than human feedback.
Comparing teacher E-feedback, AI feedback, and hybrid feedback in enhancing EFL writing skills
In a randomized controlled trial of 88 Iranian EFL learners, hybrid (teacher + AI) feedback produced the largest improvements in writing, particularly in task achievement, coherence, and grammatical accuracy, while AI feedback alone was most effective for lexical resources.
AR-Enhanced AI Feedback and EFL Writing Engagement
In a quasi-experimental mixed-methods study of 48 Spanish EFL students, AR-enhanced AI feedback led to significantly higher behavioral, emotional, and cognitive engagement, as well as increased intrinsic motivation and autonomy, compared to traditional teacher feedback.
Can AI feedback stand alone for fostering students’ presentation performance? A comparison of AI-only versus teacher-supported AI feedback
In an experimental study of 60 undergraduates, AI-only feedback was as effective as teacher-supported AI feedback for improving presentation skills and reducing anxiety in an immersive virtual reality environment, with students valuing AI as an independent feedback source.