Scholarlink

<p>The advent of advanced artificial intelligence (AI)—particularly generative-AI (GAI) applications such as ChatGPT, Midjourney, DeepSeek, and DALL·E 3—has catalysed profound change in work and value creation across industries. Organisations are striving to capitalise on AI’s promise and increasingly expect AI to empower employees to engage in infusion use—yet such aspirations are not always realised. We explain why and when AI-enabled psychological empowerment translates into post-adoptive information-systems infusion—distinguishing extended, integrative, and emergent use—and test how motives condition these pathways. We argue that empowerment generally promotes deeper use, but motive heterogeneity can enable or derail this translation. We surveyed employees (<i>N</i> = 515) about using generative-AI tools in the workplace. We employed a moderated–mediation structural equation model linking AI and empowerment to extended, integrative, and emergent use, with utilitarian (UG) and hedonic (HG) gratifications as moderators to capture motivational heterogeneity. AI-empowerment showed positive effects on integrative and extended use, which both showed positive associations with emergent use. UG strengthened the pathway from AI-empowerment to emergent use (via integrative use) and the AI-empowerment to emergent use (via extended use) pathway. HG weakened the path from AI-empowerment to extended use (a negative HG × AI-empowerment interaction), suggesting that enjoyment-oriented motives may fail to trigger—and can even dampen—depth- and breadth-oriented expansion among empowered users. These results enrich theoretical understanding of users’ AI-use patterns and inform the design of AI enablement and governance: organisations should pair empowerment initiatives with utility-framed goals to drive extended and integrative use, and monitor hedonic drift, which may fragment systematic feature expansion.</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

Gratified use of AI moderating the pathway from sense of empowerment towards infusion use

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年7月20日

查看全文
  • Yingnan Shi,
  • Yifan Zhong,
  • Chenxiao Wang

摘要

The advent of advanced artificial intelligence (AI)—particularly generative-AI (GAI) applications such as ChatGPT, Midjourney, DeepSeek, and DALL·E 3—has catalysed profound change in work and value creation across industries. Organisations are striving to capitalise on AI’s promise and increasingly expect AI to empower employees to engage in infusion use—yet such aspirations are not always realised. We explain why and when AI-enabled psychological empowerment translates into post-adoptive information-systems infusion—distinguishing extended, integrative, and emergent use—and test how motives condition these pathways. We argue that empowerment generally promotes deeper use, but motive heterogeneity can enable or derail this translation. We surveyed employees (N = 515) about using generative-AI tools in the workplace. We employed a moderated–mediation structural equation model linking AI and empowerment to extended, integrative, and emergent use, with utilitarian (UG) and hedonic (HG) gratifications as moderators to capture motivational heterogeneity. AI-empowerment showed positive effects on integrative and extended use, which both showed positive associations with emergent use. UG strengthened the pathway from AI-empowerment to emergent use (via integrative use) and the AI-empowerment to emergent use (via extended use) pathway. HG weakened the path from AI-empowerment to extended use (a negative HG × AI-empowerment interaction), suggesting that enjoyment-oriented motives may fail to trigger—and can even dampen—depth- and breadth-oriented expansion among empowered users. These results enrich theoretical understanding of users’ AI-use patterns and inform the design of AI enablement and governance: organisations should pair empowerment initiatives with utility-framed goals to drive extended and integrative use, and monitor hedonic drift, which may fragment systematic feature expansion.