<![CDATA[The determinants for continuance intention to use generative artificial intelligence (GenAI) for self-directed learning activities are explored. The research incorporates the Technology Acceptance Model (TAM), adding privacy concern and perceived risk of AI hallucination as factors that hinder adoption, and attitude toward using GenAI as a mediator. The study employed a quantitative method by using an online survey that was sent to respondents who were postgraduate students and employees of government institutions involved in using GenAI for self-directed learning. The 100 valid responses were analyzed with the Partial Least Squares Structural Equation Modeling (PLS-SEM) method by using the software SmartPLS 4. The results show that attitude towards the GenAI use has a significant positive impact on continuance intention, and Privacy concern has a significant negative impact on continuance intention. The perceived risk of AI hallucination, however, has no significant impact on attitude and continuance intention. Furthermore, the results suggest that the relationship between privacy concern and continuance intention is not mediated by attitude, and the same applies for the relationship between AI hallucination risk and continuance intention. These findings indicate that the issue of data privacy outweighs technical errors or inaccuracies in the GenAI system in determining continued use for self-directed learning. The study highlights the importance of strengthening privacy protection policies and improving AI literacy to support sustainable GenAI usage in educational contexts.]]>
