Institutional support for ethical AI adoption in higher education amid the rising trend of manuscript retractions

The growing number of papers retracted for violating AI-related ethical standards has intensified uncertainty and fear among postgraduate students, particularly because publication is often one of the compulsory requirements for their graduation. This study aimed to extend the traditional Technology Acceptance Model (TAM) by incorporating three additional constructs: (i) institutional support, (ii) knowledge of AI, and (iii) confidence in accepting (adopting) the technology. It is generally hypothesized that institutional support is essential in mitigating ethical concerns as it enhances students’ knowledge and confidence, ultimately influencing AI adoption. An online questionnaire was designed and distributed to postgraduate students at four higher education institutions, yielding 772 valid responses from postgraduate students across four universities in Tanzania. The SmartPLS-SEM approach was used to analyse the data, and the results confirmed the total mediation effect of confidence in the relationship between knowledge of AI and AI adoption. These findings expand the traditional TAM, increase its explanatory power, and underscore the significance of higher education institutional support in fostering AI literacy and confidence among students, ensuring ethical AI integration in their professional lives.