GenAI as a frenemy in teaching: Perceived autonomy and risks
DOI:
https://doi.org/10.14742/ajet.11301Keywords:
artificial autonomy, human-AI interaction, GenAI adoption, risk aversion, mixed methodAbstract
Generative artificial intelligence (GenAI) has the potential to transform education, yet little is known about how teachers’ perceptions of its autonomy and their risk aversion influence their willingness to adopt. This mixed method study utilised a human-AI interaction lens to investigate the psychological processes shaping varied adoption. Survey data from 287 teachers across 27 countries and regions were analysed using structural equation modelling with bootstrapped mediation tests, complemented by thematic analysis. Research results revealed that integrating GenAI in teaching is neither straightforward nor effortless. Teachers largely viewed it as semi-autonomous, preferring to retain full control over their teaching. Their awareness of situational factors may contribute to the perception of GenAI as semi-autonomous and weaken the link between artificial autonomy and their adoption intention. As the appeal of automation diminished, concerns about potential harm grew, with risk aversion playing a significant role in adoption decisions. Teachers were more concerned about students’ use of GenAI than their own. Overall, GenAI adoption requires careful consideration of context and human-AI interaction. This study enhances understanding of these dynamics, highlighting the psychological processes influencing adoption and offering insights for AI policy and professional training aimed at fostering thoughtful integration into education.
Implications for practice or policy:
- Instructional designers could improve adoption by sharing discipline-specific examples aligned with instructional needs.
- Institutions could tailor training to teachers’ GenAI use experience, with novices introduced to pedagogical potential and experienced teachers supported with practical strategies.
- Institutions could fund AI literacy initiatives, integrating AI ethics into curricula and adapting assessments for ethical and effective AI use.
- Policymakers should develop clear guidelines and risk management strategies to address integrity and student misuse.
Downloads
Metrics
Downloads
Published
How to Cite
Issue
Section
License
Copyright (c) 2026 Hai Min Dai, Kaige Ni, Han Meng, Bei Ju, Joseph Crawford, Timothy Teo

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.
Articles published in the Australasian Journal of Educational Technology (AJET) are available under Creative Commons Attribution Non-Commercial No Derivatives Licence (CC BY-NC-ND 4.0). Authors retain copyright in their work and grant AJET right of first publication under CC BY-NC-ND 4.0.
This copyright notice applies to articles published in AJET volumes 36 onwards. Please read about the copyright notices for previous volumes under Journal History.
