01641nas a2200145   4500000000100000000000100001008004100002100001800043700001400061700001400075700001700089245004400106856011400150520123100264       2024                            d1 aGuoliang Zhou1 aYijia Liu1 aZheng Yan1 aErol Gelenbe00aIs ChatGPT Trustworthy Enough? A Review  uhttps://www.researchgate.net/publication/385065063_Is_ChatGPT_Trustworthy_Enough_A_Review#fullTextFileContent3 a<p>ChatGPT, as an advanced model that seamlessly integrates into diverse digital interactions,<br />
shows great potential to enhance the performance of consumer technology and reshape its<br />
landscape. The critical question of its trustworthiness and associated challenges becomes<br />
increasingly prominent. However, the literature still lacks a thorough review to study its trust. This<br />
comprehensive review explores whether ChatGPT is trustworthy enough by navigating the<br />
multifaceted realm of large language models, placing a significant focus on its exemplar model,<br />
ChatGPT. Our exploration traverses the complex interplay of factors impacting user trust, including<br />
trust in ChatGPT itself and trust in its utilization and dissemination. By delving into insightful<br />
perspectives on the ChatGPT trustworthiness regarding a set of evaluation criteria including<br />
fundamental properties, subjective properties, security and privacy, we find that ChatGPT is far from<br />
trustworthy based on related literature review. This paper sheds light on the weakness of ChatGPT<br />
trust and provides the trajectory of future development in the realm of large language models</p>