Seeing Ourselves: Bioethical Frameworks for AI Body Visualization in Gender-Affirming Care

Abstract Book of the 5th International Conference on LGBT studies

Year: 2026

[PDF]

Seeing Ourselves: Bioethical Frameworks for AI Body Visualization in Gender-Affirming Care

Nadia Mariam Mazloum

ABSTRACT:

There is a lack of evidence about using AI algorithms to ethically AI to train data models for representing gender-affirming care. New research done by Piluso and Buslón documents both the therapeutic value of existing consumer tools and their serious failures, including binary bias, western-centric training data, and lack of clinical guidance. The primary counterargument, psychological harm from unrealistic expectations, is addressed as a design problem, not grounds for banning the algoriths. Rather, the ethical question is not whether this technology should exist, but under what conditions it can be made genuinely beneficial. The current paper proposes a discussion on the ethics of training data ethics: the consent of data subjects, fair burden of data collection, and structural inadequacy of binary datasets. Beauchamp and Childress’s four-principle framework, the analysis shows that properly governed tools promote autonomy, advance beneficence, and serve justice. Responsibility requires not only clinical governance and privacy protections but community co-design with Trans, Non-binary, and Cis-gender individuals at every development stage. Concurrent empirical research examines actual user perspectives through a survey investigating experiences with existing tools, privacy preferences, training data ethics, and willingness to participate in tool design. This paper argues that AI-assisted body representation ools can be ethically justified in gender-affirming care when developed with proper privacy protections and clinical oversight.

Keywords: Artificial Intelligence; Bioethics; Community Co-Design; Gender-Affirming Care; Health Equity