B.C.’s privacy watchdog weighs in on health AI boom – as doctors warn it’s not a substitute

In all likelihood, your last hospital or doctor visit involved artificial intelligence tools.

As a growing number of doctors adopt artificial intelligence tools in their offices and hospitals, British Columbia’s privacy commissioner is urging them to do their homework on privacy requirements.

CTV News sat with Michael Harvey, B.C.’s Information and Privacy Commissioner, for an in-depth discussion around AI and found the area isn’t just nuanced, it’s being developed and assessed as the technology evolves.

“There is no question that our laws need to be reformed to adapt to the changing technological circumstances,” he stated. “That said, it’s not like there’s no laws that apply to AI in the health sector or other sectors.”

Notifying patients they’re using the technology is the minimum, Harvey said, but health-care practitioners should go further if they’re using scribes or other software in their practice.

“Even in situations where it might not be strictly legally necessary to do more than notify, I think it’s a good advice for clinicians to really take that extra step and have a bit of a conversation,” said Harvey. “Because we’re talking about new types of applications here, organizations would be well advised to hold themselves to a higher bar of express consent.”

The Doctors of BC has established a working group to draft a policy framework for their members and vet AI software that meets ethical and legal requirements in the province.

What’s clear and not so clear

There are two aspects privacy watchdogs are monitoring as AI permeates the health-care system, in particular: the data being used to train the models, and the experience of the patients receiving care.

Harvey said if a provider is gathering information they need to notify the patient, and if it’s being used for a secondary purpose like training an AI, “generally speaking, you should have to consent for that purpose, but there are exceptions in the law.”

He was clear that while a program could be approved for one type of use, it can’t just be used for another purpose without a fresh assessment of the privacy impact “because sometimes that can even change the whole legal basis for the program,” and whether it’s in compliance with B.C. privacy laws, which aren’t the same as our federal legislation or U.S. HIPAA laws.

Harvey encouraged patients to ask their health-care provider questions about what kind of data is being used, and to contact his office to report any red flags that he may need to take a look at, since “protecting people’s trust in the health system should be a very high priority.”

Doctors warn of limits

As the province’s health-care system continues to see long waits for patients to see family doctors or emergency room physicians, Doctors of B.C. worries that patients will try to self-diagnose with AI when the technology isn’t meant for that purpose.

“We don’t want serious illnesses or serious conditions to be missed,” said president Dr. Charlene Lui, who insisted that the public and health-care providers alike should consider AI to be a tool, rather than a substitute.

She said that every doctor has stories of experiences throughout their career where they spotted signs of cancer, diabetes, or other serious medical conditions while meeting with a doctor for another reason.

Lui is a family doctor herself, and will always remember an appointment with a woman who’d gone to see her for her own health issue and had brought her baby – who immediately caught Lui’s attention and was quickly rushed to hospital.

“The baby had heart surgery that day,” she said. “There is something about seeing a physician that quick scan that a physician does that I think is often underappreciated.”

This is the second part in a CTV Vancouver series taking a deep dive into the use of artificial intelligence in health care. You can read part one on everyday uses here.