When Missing Data Become Missing Rights: AI and Women’s Health in Crisis and Post-Conflict Settings
Abstract accepted for presentation · Acceptance received October 8, 2026
Nazanin Azizmohammadi, Ph.D.
Status: Abstract accepted for presentation; acceptance received on 8 October 2026.
Congress: UNESCO Chair on Health, Race and Human Rights Global Congress 2027.
Theme: Truth-Telling, Human Rights, and Health Justice in an Age of Global Fracture.
Organiser: UNESCO Chair on Health, Race and Human Rights, Peter A. Allard School of Law, University of British Columbia.
Congress dates and location: 1–3 February 2027, San José, Costa Rica.
Language: English.
Abstract
Artificial intelligence increasingly informs health prioritisation and resource allocation. In crisis and post-conflict settings, however, the records used by these systems may exclude women whose access to care is constrained by displacement, poverty, disability, insecurity, or fear of disclosing sexual violence. When recorded utilisation becomes a proxy for medical need, absence from data can be mistaken for absence of need.
This paper examines when reliance on incomplete health data becomes a failure of legal accountability. Using doctrinal and normative analysis, it brings the right to health and substantive equality into dialogue with intersectional feminism and epistemic injustice. Its legal framework draws on the International Covenant on Economic, Social and Cultural Rights, General Comments 14 and 22, and WHO and UNESCO guidance on artificial intelligence.
Through illustrative scenarios concerning maternal care, mental health after violence, and treatment allocation, the paper argues that aggregate accuracy and nominal human oversight cannot establish the legitimacy of algorithmic decisions. Institutions must justify how needs are defined, whose experiences are omitted, and how uncertainty and error affect differently situated women.
The paper proposes a gender-responsive right-to-health impact assessment linking data representation, scrutiny of proxies, disaggregated performance, privacy, meaningful participation, human review, and accessible remedies. Emergency deployment should remain time-limited and subject to reassessment. The central claim is that institutions must demonstrate adequacy of their evidence rather than require excluded women to prove needs that existing records fail to capture.