Research on Trustworthiness Analysis Technology of Artificial Intelligence Software
Shen Xiaomei, Wu Lijin, Jin Kuangyu, Han Xinyu, Longli Tang · 2023
In order to solve the trustworthiness risk problems of artificial intelligence software such as algorithm model security, data security and privacy risk, infrastructure security and other types of trustworthiness risks, we investigate the trustworthiness mechanism and defect patterns of artificial intelligence software and proposes a comprehensive analysis method of artificial intelligence software trustworthiness based on SFMEA and SFTA, which realizes a comprehensive analysis of artificial intelligence software trustworthiness. Through example analysis, this method can effectively improve the quality level of artificial intelligence software.