Scholarlink

<p>The proliferation of artificial intelligence (AI) in academic writing raises critical questions about institutional legitimacy and professional identity within scholarly communities. This study develops and empirically tests the AI Authorship Legitimacy Model (AALM), a novel theoretical framework that examines how perceived AI authorship capability is associated with institutional legitimacy through the mediating role of academic identity threat, moderated by governance mechanisms. Using structural equation modelling (SEM) with data from 384 academic journal editors and reviewers across multiple disciplines, we find that perceived AI authorship capability is positively associated with institutional legitimacy (β = 0.36, <i>p</i> &lt; .001) whilst simultaneously associated with lower academic identity threat (β = −0.31, <i>p</i> &lt; .001). Academic identity threat is negatively associated with institutional legitimacy (β = −0.42, <i>p</i> &lt; .001). Contrary to conventional wisdom, strong governance mechanisms amplify rather than constrain the positive relationship between AI capability and legitimacy (β = 0.18, <i>p</i> = .003). The model explains 48% of variance in institutional legitimacy, suggesting robust predictive validity. These findings extend institutional theory by identifying identity threat as a critical psychological mechanism underlying technological legitimation, whilst providing actionable guidance for research institutions developing AI governance frameworks.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Governance and legitimacy in AI: developing and testing the AI Authorship Legitimacy Model (AALM)

  • 期刊论文

  • 发表日期：

    2026年5月4日

查看全文
  • Sam Adeyemi

摘要

The proliferation of artificial intelligence (AI) in academic writing raises critical questions about institutional legitimacy and professional identity within scholarly communities. This study develops and empirically tests the AI Authorship Legitimacy Model (AALM), a novel theoretical framework that examines how perceived AI authorship capability is associated with institutional legitimacy through the mediating role of academic identity threat, moderated by governance mechanisms. Using structural equation modelling (SEM) with data from 384 academic journal editors and reviewers across multiple disciplines, we find that perceived AI authorship capability is positively associated with institutional legitimacy (β = 0.36, p < .001) whilst simultaneously associated with lower academic identity threat (β = −0.31, p < .001). Academic identity threat is negatively associated with institutional legitimacy (β = −0.42, p < .001). Contrary to conventional wisdom, strong governance mechanisms amplify rather than constrain the positive relationship between AI capability and legitimacy (β = 0.18, p = .003). The model explains 48% of variance in institutional legitimacy, suggesting robust predictive validity. These findings extend institutional theory by identifying identity threat as a critical psychological mechanism underlying technological legitimation, whilst providing actionable guidance for research institutions developing AI governance frameworks.