NT Health (Royal Darwin Hospital)
No attacker, no malware, no intrusion. The same inventory problem anyway.
What happened?
Photos and videos posted to an Instagram account with 28,000 followers exposed confidential medical information belonging to pregnant patients at Royal Darwin Hospital. The material was posted by Neda Mogharehabed, an overseas trained doctor who was visiting the hospital from 12 to 14 October 2026.
NT Health initially said five patients were affected. On 28 October 2026 the department revised that figure, admitting twelve pregnant patients were impacted. The hospital terminated the doctor's attachment, issued formal apologies, and referred the matter to the public information commissioner and the police commissioner to determine whether any criminal offences occurred. Overseas doctor training placements were paused.
What data was actually inside?
Patient names, hospital record numbers, ward and bed assignments, consultant names, presenting problems, medical histories, investigations and care plans.
This is a more complete clinical picture than most ransomware incidents produce. A ransomware operator takes a database export and gets structured fields. What was published here is the working record: the presenting problem, what was investigated, what the plan is, who the consultant is, and which bed the patient is in. Ward and bed assignment is also a physical location in real time, which is a category of exposure that database breaches almost never create.
Who gets hurt and how?
Twelve pregnant patients in a hospital in a city of roughly 150,000 people.
Scale works in reverse here. In Darwin, a name attached to a ward, a consultant and a presenting problem is not anonymous in any practical sense, and obstetric care covers circumstances that patients routinely choose not to disclose to their own families. One affected patient described it as a betrayal, which is the correct word for a disclosure that came from inside the care relationship rather than from an attacker.
There is no technical remedy available. Credit monitoring is irrelevant, a password cannot be changed, and the information was accurate, clinical and published to an audience of 28,000.
What did they think they were doing right?
Supervised observation is how clinical training works everywhere. A visiting doctor on a short attachment, accompanied on a ward, is the normal mechanism by which medicine is taught, and it requires genuine access to real patients.
The hospital has since attributed the failure to training and safeguards around social media use and patient confidentiality obligations. That is a fair reading. It is also worth noting what no access control would have prevented: the doctor was authorised to be present, authorised to see the patients, and authorised to view the records. Nothing was bypassed.
What did they not know about their own data?
How many patients were in the material. The department said five and then said twelve, which is a revision of more than double, published two weeks after a three day visit.
That correction is the entire reason this incident belongs in a series about data inventory. There was no intrusion to scope, no logs to correlate, no exfiltration to measure. The source material was a finite set of posts on one public account, and the task was to look at the images and count the patients visible in them.
It still took two weeks and still produced a number that had to be revised upward. If determining who was exposed is hard when the evidence is a public Instagram account, consider how hard it is when the evidence is six terabytes taken from a file server. The Photon Health entry in this series makes the same point about a data review that took weeks to confirm its categories.
The underlying gap is that a patient record is not a unit. It is a name, a number, a bed, a history and a plan, scattered across a chart, and counting how many patients appear in a photograph requires reading all of it.
If you handle patient data, could you identify within 24 hours exactly which records were accessed in a breach?
Not sure where to start? How to find sensitive data across your estate
What does attribution look like the morning after?
Attribution was never in question, which makes this a clean experiment. The actor is identified, the mechanism is known, the material is public, and the window is three days.
The matter sits with the public information commissioner and the police commissioner. NT Health must notify each affected patient under the Information Act, and the revision means some patients learned later than others that they were among the group. Pausing overseas doctor training placements is the institutional consequence, and it will be felt by clinicians who did nothing wrong, in a health service that depends on them.
What would have changed the outcome?
Prevention here is a training and conduct question rather than a data discovery one, and it would be dishonest to pretend otherwise. Access control cannot stop an authorised clinician from photographing a screen.
The response is a different matter. The two week path from five patients to twelve is the recognisable shape of an organisation establishing after the fact what its records contain and who they belong to. Knowing which patients were on a ward during a defined window, and what identifying detail each chart carries, is answerable in advance. The institutions that notify quickly and accurately are the ones that did not have to assemble that picture while apologising for it.
Sources
NT Health (Royal Darwin Hospital) found out the hard way.
Your team could spend the next 6 months rebuilding systems, notifying customers, and answering legal questions. Or you could spend 24 hours finding out what's actually at risk.
Want to understand the process first? Read our guide to sensitive data discovery .