Patients should have the right to refuse AI scribes 'without retaliation', privacy group says

Patients should have a specific legal right to refuse the use of AI scribes during consultations, a privacy advocacy group says.

AI scribes have been one of the most quickly adopted technological changes in general practice.

But Digital Rights Watch claims GP practices are regularly failing to get informed consent before using AI scribes — a complaint also made by patient groups — which could breach privacy laws.

In a 37-page report published last week, it said practices often sought consent by doctors verbally asking, ‘Are you okay for me to use an AI scribe?’ or through signs in waiting rooms.

However, it argued these failed to meet the requirements for collecting health information under the Privacy Act 1988.

The Office of the Australian Information Commissioner says consent for collecting personal information must be adequately informed and voluntary.

“There are growing concerns that many patients are not being adequately informed of the use of AI scribes, nor is genuine consent consistently obtained, despite the requirements under the Australian Privacy Act 1988,” the report said.

The group cited an anonymous story from a father who said he had only learnt that his young son’s consultation was being scribed with AI mid-consult and the GP had been unable to provide any information about the specific software.

It also pointed to a Guardian report that a Melbourne psychiatrist had told patients who did not consent to the use of an AI scribe they would have to find another doctor.

In a separate petition for which it is seeking signatures, the not-for-profit said, “Clinics are installing AI scribes all over our health system.

“But AI scribes come with significant risks: risks to patient privacy, risks to the quality of healthcare and risks to doctors’ legal liability.

“It is entirely reasonable for a person to reject the use of AI scribes in the course of their healthcare.

“But patients are being refused healthcare for precisely this reason.”

It called for a legally enshrined right for patients to refuse the use of AI systems in healthcare “without retaliation”.

The report urged the TGA — which is currently reviewing whether AI scribes should be regulated as medical devices — to go down this path.

While most AI scribe makers say audio is deleted immediately after being transcribed, Digital Rights Watch said only Heidi Health responded to its questions on whether its scribe used consultation data to fine-tune large language models or for other commercial uses, saying it did not.

Earlier this year, the Consumers Health Forum of Australia called on the Federal Government to set minimum standards for how GPs obtained and recorded patient consent for AI scribe use.

Its CEO, Dr Elizabeth Deveny (PhD), said a quick verbal consent when a consultation started would not be enough, given patients may have questions but not want to waste consultation time asking them.

More information: The Guardian; 19 May 2026.

You're missing out on this article.