Canadians and AI: A Complicated Health Care Relationship
Canadians are no strangers to typing symptoms into a search bar at 2 a.m. — and increasingly, that search is landing on an AI chatbot. But according to a new national poll, comfort with AI in health care drops sharply the moment those tools start making actual medical decisions.
The survey, released this week, found that while a notable share of Canadians have already used AI-powered tools like ChatGPT or similar platforms to get medical information or advice, most draw a firm line at replacing the human side of medicine. Respondents were significantly less comfortable with AI diagnosing conditions, recommending treatments, or standing in for an in-person appointment with a physician.
What the Poll Actually Found
The results paint a picture of cautious adoption. Many Canadians see AI as a useful first step — a way to understand symptoms, research conditions, or prep questions before a doctor's visit. Think of it as the 21st-century version of flipping through a medical encyclopedia, but faster and more conversational.
Where trust drops off is in higher-stakes scenarios. Survey respondents showed significantly lower comfort levels when asked about AI systems making treatment decisions or being used as a substitute for a regulated health care professional. The message from Canadians seems to be: helpful tool, yes — but keep a human in the loop.
Why This Matters Now
The poll lands at a moment when Canada's health care system is under considerable strain. Family doctor shortages are a reality in communities from British Columbia to Newfoundland, with millions of Canadians currently without a family physician. Wait times for specialists remain long in many provinces.
In that context, it's not hard to see why AI tools are filling a gap — at least informally. When you can't get an appointment for three weeks, asking an AI chatbot whether your symptoms warrant urgent attention feels less like a novelty and more like a necessity.
But health care experts have been quick to flag the risks. AI models can hallucinate, produce outdated information, or fail to account for a patient's full medical history. What feels like reliable advice from a chatbot may be confidently wrong — and in health care, that gap between confidence and accuracy can be dangerous.
The Bigger Question for Canadian Health Policy
The poll raises questions that go beyond individual Canadians and their search habits. Health technology companies are actively pushing AI tools into clinical settings — from diagnostic imaging assistance to triage chatbots — and regulators are scrambling to keep pace.
Health Canada has signalled it's paying attention to AI in medical devices and clinical decision support, but a comprehensive national framework for AI in health care is still a work in progress. The public sentiment captured in this poll may end up being an important data point for policymakers deciding how quickly — and how far — to let AI into the exam room.
For now, most Canadians seem to have landed on a pragmatic middle ground: use the tools that are available, but don't let an algorithm be the last word on your health.
Source: CBC News. Original reporting via CBC Top Stories RSS feed.
