Choices of Artificial Intelligence (AI): ChatGPT's Solutions to Ethical Dilemmas in Bipolar Disorder Care

Russell Franco D'Souza; Krishna Mohan Surapaneni; Mary Mathew; Shabbir Amanullah; Rajiv Tandon

doi:10.1111/bdi.70059

Choices of Artificial Intelligence (AI): ChatGPT's Solutions to Ethical Dilemmas in Bipolar Disorder Care

Bipolar Disord. 2025 Nov;27(7):511-518. doi: 10.1111/bdi.70059. Epub 2025 Sep 1.

Authors

Russell Franco D'Souza¹, Krishna Mohan Surapaneni^{2

3}, Mary Mathew⁴, Shabbir Amanullah⁵, Rajiv Tandon⁶

Affiliations

¹ Organizational Psychological Medicine, International Institute of Organisational Psychological Medicine, Melbourne, Australia.
² Department of Biochemistry, Panimalar Medical College Hospital & Research Institute, Chennai, Tamil Nadu, India.
³ Department of Medical Education, Panimalar Medical College Hospital & Research Institute, Chennai, Tamil Nadu, India.
⁴ Department of Pathology, Kasturba Medical College, Manipal, Manipal Academy of Higher Education, Manipal, Karnataka, India.
⁵ Division of Geriatric Psychiatry, Queen's University, Kingston, Ontario, Canada.
⁶ Department of Psychiatry, WMU Homer Stryker School of Medicine, Kalamazoo, Michigan, USA.

PMID: 40891014
DOI: 10.1111/bdi.70059

Abstract

Background: Bipolar disorders present complex ethical challenges to patient care due to the delicate balance between patient autonomy and safety. The use of artificial intelligence (AI), particularly ChatGPT, holds the potential to address these dilemmas by providing personalized treatment plans, monitoring patient well-being, and reducing stigma associated with mental health issues. However, the application of AI in this context requires a deep understanding of the unique needs and vulnerabilities of individuals with bipolar disorder.

Methods: This experimental study evaluated ChatGPT's (Version 3.5) responses to ethical dilemmas in bipolar disorder care using three clinical case scenarios reported in an open-access publication. The study compared ChatGPT's answers to an original answer key from the article. ChatGPT's responses were cross-checked and analyzed for alignment with the explanation given in the article.

Results: ChatGPT provided mostly congruent responses with the original answer key, demonstrating its potential to offer insights and considerations for ethical dilemmas. However, there were variations in some responses, emphasizing the complexity of ethical decision-making in healthcare. These findings underscore the importance of combining AI-generated insights with human expertise in complex medical and ethical situations.

Conclusion: ChatGPT, and similar AI systems, can be valuable resources for addressing ethical concerns in bipolar disorder care. They offer guidance and information to clinicians, patients, and stakeholders, contributing to shared decision-making in healthcare. Nonetheless, the study highlights the essential role of human judgment and expertise in navigating intricate ethical dilemmas. Continuous research and development are necessary to enhance ChatGPT's capabilities and ensure responsible use, aligning AI assistance with the highest standards of patient care and ethical conduct.

Keywords: ChatGPT; artificial intelligence; bipolar disorders; ethical dilemmas.

Publication types

Case Reports

MeSH terms

Adult
Artificial Intelligence* / ethics
Bipolar Disorder* / therapy
Decision Making*
Ethical Dilemmas
Female
Generative Artificial Intelligence
Humans
Male
Middle Aged