Abstract: [Purpose/Significance] With the rapid development and widespread application of artificial general intelligence（AGI） technology represented by generative artificial intelligence, the "AI divide" has emerged. A comprehensive portrayal of the AI divide can provide valuable reference for policy-making aimed at bridging the gap and enhancing the inclusiveness in the digital society. [Design/Methodology] This study used semi-structured interviews to collect raw data from individual cases. Based on the interview data, a preliminary AI divide scale was developed through systematic coding. Then, based on the pre-survey results of the scale, a questionnaire survey was conducted for both parents and offspring. Finally, a formal scale was revised and formed. [Findings/Conclusion] After verification through methods such as Cronbach's α coefficient test, KMO test, and Bartlett's sphericity test, an AI divide measurement system applicable to intergenerational contexts, which includes four dimensions: access divide, psychological divide, literacy divide, and benefit divide is constructed. The results of the scale application show that there is a significant intergenerational AI divide, with the offspring outperforming parents in the access divide, literacy divide and overall dimension. [Originality/Value] This study explores the measurement theories and methods of the generational digital divide in the AI era, enriching the theoretical framework on the relationship between AI and society.

Keywords: Artificial general intelligence（AGI）, AI divide, Digital divide, Scale development