Abstract
Due to sanctions, more Chinese high-tech SMEs are turning to rent AI computing power through cloud service providers. Therefore, it is necessary to give a variety of suggestions for China's high-tech SMEs to better develop AI applications through computing power leasing. Because traditional theories are difficult to explain this new technology adoption behavior, this research combines and extends TTF and UTAUT2 theories to take an empirical research. A total of 387 questionnaires were received, of which incomplete questionnaires and invalid questionnaires were issued, leaving 281 valid questionnaires. The results indicate that SME innovativeness, perceived risk, performance expectancy, price value and task technology fit are all significantly related to usage, whereas task technology fit moderates the other relationships significantly. Results give a variety of suggestions for China's high-tech SMEs to better develop AI applications through computing power leasing in the context of sanctions. This study not only suggests ways to increase the competitiveness of SMEs by optimizing leasing services but also give directions in investors' investment decisions. The findings are also applicable to the large-scale application of China's domestic AI chips in computing power leasing scenarios in the future.