Artificial intelligence (AI) is often proposed as a solution to rural and remote health workforce shortages. However, when AI systems are trained on historical service utilisation data shaped by cost, cultural, and logistical barriers, they risk misinterpreting constrained access as low demand and perpetuating inequities rather than reducing them. This research monograph quantifies inequities in (i) health workforce distribution, (ii) digital infrastructure, and (iii) health service utilisation for Aboriginal and Torres Strait Islander peoples and remote communities, using Australian government datasets spanning 2013–2025. Findings identify a 72% specialist workforce deficit in remote areas, a 21‑point digital inclusion gap in Very Remote communities, and significant access barriers including cost (41.6%), cultural safety concerns (25.6%), and logistical constraints (24.9%). Combined through a multiplicative model, these inequities reduce effective access for Indigenous Very Remote populations to approximately 12% of the urban non‑Indigenous baseline. The analysis demonstrates how AI systems can encode “prophetic surveillance”—forecasting the continuation of structural barriers—and reveals how representation bias, proxy variables, and infrastructure mismatches produce high‑risk deployment environments. This monograph provides an empirical foundation for AI governance in Australian health systems, offering a structured framework for evaluating equity risks, preventing algorithmic harm, and prioritising community‑led models and structural reforms over technologically driven substitutions.
Stephanie Fletcher (Fri,) studied this question.