AI enabled healthcare
Governing the afterlife of patient data
Article | 2026-08-26
5 minute read
Australian health services are being asked to share more information to improve care quality. The harder question is what AI can infer from routine clinical and administrative data. For leaders, the issue is no longer only what a record contains. It is what that record enables a system to conclude.
Health services need to govern consequential conclusions derived from health information, not only the information itself. Permission drift occurs when authority stretches into new purposes, organisations or technologies, or remains in use after circumstances change. Inference risk arises when analytics or AI creates high-impact information the person never disclosed. Privacy governance follows source data, and clinical governance follows final decisions, but the inference between them can remain active without clear accountability.
The problem becomes visible when a patient withdraws consent, changes an access preference or corrects source information. Future use may stop, yet a risk score may remain in the clinical record, or the patient’s data may already have shaped a trained model. Each needs a different governance response.
Health leaders therefore need inference stewardship. Every material inference should have an owner, defined purpose, evidence limits, review triggers and a path for challenge, correction, restriction or retirement. Responsible information sharing should be judged by whether the health service remains accountable for the conclusions it uses.
Permission can drift forwards and persist backwards
The case for information sharing is strong. Clinicians need timely access to pathology, imaging, medicines and discharge information. Consumers should not have to repeat their history or undergo duplicate investigations. The Australian Government’s Sharing by Default impact analysis identified incomplete information as a constraint on participation, decision making and coordination (*1).
From 1 July 2026, specified pathology and diagnostic imaging reports must generally be uploaded to My Health Record unless an exception or extension applies. Consumers can request that reports not be uploaded, and later delete, hide or restrict access to reports already present.
Those controls matter, but their reach is limited. They apply to My Health Record, not necessarily to provider systems, copied information, downstream conclusions or trained models. My Health Record is also not a complete health record (*2).
Permission drift begins when authority granted for one context is carried into another, such as service improvement, research, model development or supplier reuse. Each approval may appear reasonable, while no single forum sees the full chain from original understanding to downstream effect.
A 2025 scoping review identified inadequate consent, privacy and security concerns, and unauthorised sharing as barriers to using health information in AI models. Better consent, stronger data protection and ethical standards were key facilitators (*3). Different purposes need defensible authority and a clear explanation.
Permission can also drift over time. A conclusion can continue to influence care after consent is withdrawn, information is corrected, evidence evolves or the approved purpose ends. Australian privacy guidance states that consent must be current and specific, may be withdrawn, and cannot support future use or disclosure after withdrawal (*4).
Health services need to know the basis for each use, whether it still applies and how existing outputs should be handled. Consent, care obligations, legislation, expectations and research arrangements can each lead to a different answer. Continuation should be reviewed, not assumed.
Where care information is later used for model development, consent to receive a service should not be treated as consent for AI training. Leaders should separate authority for care, secondary use and model development, then test whether downstream analytics can still infer what the patient sought to limit (*5).
