Edited by: Peter Kokol, University of Maribor, Slovenia
Reviewed by: Leila Allahqoli, Iran University of Medical Sciences, Iran
*Correspondence: Wenting Tong,
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
Joseph Kassab’s recent article (
ChatGPT is a powerful language model developed by OpenAI based on the GPT-3/GPT-4 architecture that has demonstrated remarkable accuracy and broad potential for application in medical diagnosis. Although some research suggests that ChatGPT needs to be specifically trained in the medical field to be effective (
Collaboration with medical experts is essential to regularly assess and validate its diagnostic suggestions to ensure accuracy and reliability. Additionally, since ChatGPT cannot provide medical diagnoses or treatment plans autonomously, such systems cannot be licensed for medical practice.
Another challenge is privacy concerns. Disputes over privacy data on a global scale have led to the prohibition of ChatGPT usage in some countries, countries such as Russia, Italy, and China have either banned or previously imposed bans on the use of ChatGPT. To overcome an issue that may be more political than scientific or privacy-related, stricter data protection policies and regulations could be established to ensure data is utilized in compliance with laws while safeguarding user privacy.
Moreover, limited access to patient data and language bias are other challenges that need to be addressed. Also, due to limitations in training data, ChatGPT exhibits a significant performance gap between English and non-English scenarios, preventing non-English speakers from fully benefiting from its resources. To resolve these issues, the collection and training of non-English data should be expanded to enhance AI performance across various language environments.
Furthermore, OpenAI has restricted access to ChatGPT for users in some countries, leading to regional bias. Collaboration with relevant countries and regions should be pursued to jointly explore potential technology applications and proactively expand the scope of services, ensuring that users worldwide can benefit. Fee issues also need to be addressed, as the fee-based model may lead to accessibility imbalances among countries with different income levels. Flexible pricing strategies can be adopted, offering various pricing schemes based on national income levels and user needs, ensuring that everyone can enjoy the conveniences brought by AI.
Finally, legal liability and sensitivity to problem nuances are other challenges that need to be considered. Further refinement of the legal system is needed to clarify the legal responsibility for AI decision-making. Subtle differences between questions may yield different answers, and continuous optimization of AI algorithms is required to improve the understanding and accuracy of responses.
Although AI technologies like ChatGPT face numerous challenges, addressing these issues requires a multifaceted approach, incorporating technological innovation, ethics, law, data security, and education. The advancement of AI is inevitable, and it is essential to adopt a proactive and flexible mindset towards tackling these challenges. Governments, businesses, research institutions, and the public should collaborate, share resources and knowledge, and expand cooperation to optimize AI technology and improve AI applications in healthcare. Additionally, it is important to address other elements influencing healthcare resource distribution, such as infrastructure development, medical personnel training, and public health policies. AI technology can complement and enhance traditional healthcare systems, but it cannot replace them entirely. Focusing on resolving issues within conventional healthcare systems is also crucial to deliver more comprehensive, efficient, and sustainable medical services to the public.
HZ led the conceptualization of the study, literature review, and initial drafting. YG provided insights into AI technology in healthcare, including challenges and potential solutions, and contributed to manuscript revision. JC examined privacy and data security issues, also contributing to manuscript development. WT explored socio-political implications and discussed disparities in AI accessibility, participating in manuscript writing and editing. All authors approved the final version and agreed to be accountable for their own contributions.
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.