How One Ontario Clinic Told Patients About AI Before They Had to Ask
AI transparency is becoming one of the most important issues facing Ontario healthcare organizations. Most healthcare organizations adopt new technology quietly. It gets installed, staff get a memo, and patients find out when something feels different about a phone call or an appointment reminder.
One Ontario family clinic decided to do the opposite.
Before rolling out any AI-assisted tools, the clinic published a plain-language page on its website explaining exactly what AI would do, what it would never do, and how patients could always reach a human being. The page opens with a line that should be standard practice across healthcare:
“We want you to hear this from us first, not find out later.”
That single sentence does more for patient trust than any feature list ever could.
What the clinic got right
The page is worth studying because it answers the three questions every patient actually has about AI in a medical setting.
What does it do here? The clinic listed specific, limited uses: organizing administrative information, appointment reminders and recall notifications, and supporting documentation workflows. No vague promises about “innovation.” Just the real, boring, useful work.
What will it never do? This is the section most organizations skip, and it is the one patients care about most. The clinic stated plainly that AI does not diagnose conditions, does not make treatment decisions, and does not replace a conversation with a doctor. AI supports human decision-making. It never replaces it.
How is my information protected? As an Ontario health clinic, the page grounds everything in the Personal Health Information Protection Act, 2004 (PHIPA). It commits to collecting only the minimum information necessary, reviewing every AI tool for privacy impact before it is introduced, and holding any technology partner to the same obligations the clinic itself carries under a written agreement. It also makes a point many organizations miss: anything an AI tool produces, whether a note, a summary, or a message, is protected under Ontario privacy law the same way a paper record is.
The rule that should become industry standard
Buried in the middle of the page is the most important commitment:
If an AI tool is ever part of a phone call or message, it will identify itself. And patients can always reach a person.
At Ajax Web AI, this is not optional. Every AI voice agent we build discloses that it is AI, in every industry, for every client. A human handoff is always available. We believe disclosure builds trust faster than a perfect imitation ever could, and trust is the entire point.
If your AI receptionist has to pretend to be human to work, it does not work.
Why transparency is a competitive advantage, not a compliance chore
Here is what most clinics get wrong about AI adoption. They treat patient communication about AI as a risk to manage. Something legal reviews, something you publish only if someone complains.
The clinic in this example treated it as a service to patients. And the result is a page that does three jobs at once:
- It answers questions before they become complaints. A patient who reads “you can always ask for a person” is not going to feel tricked by an automated reminder call.
- It demonstrates PHIPA accountability publicly. Regulators, families, and community partners can see the clinic’s obligations spelled out in plain language.
- It positions the clinic as a leader. Most healthcare organizations in Ontario have not published anything like this. The ones that do first will own the trust conversation in their communities.
Transparency is one piece of a larger toolkit
A transparency page is where responsible AI adoption becomes visible to patients. But it is the last step, not the first. Behind a page like this sits a set of decisions every healthcare organization needs to make before an AI tool touches a single phone call:
- A privacy review completed before launch, not after
- A written agreement holding the technology partner to PHIPA obligations
- A clear human escalation path that staff actually know how to use
- Staff who can answer patient questions about the AI without guessing
- An AI disclosure policy applied consistently, every call, every message
- A schedule for reviewing how the tools are performing and whether the page still reflects reality
This is why we tell clinics we do not sell an AI receptionist. We help organizations adopt AI in a way that patients, families, staff, and regulators can trust. The voice agent is one component. The confidence around it is the product.
What your clinic can do this month
If you are a clinic owner or manager in Durham Region or anywhere in Ontario thinking about AI tools, start here:
- Write down what AI will and will not do in your practice. If you cannot state it plainly, you are not ready to deploy it.
- Confirm your vendor operates under a written PHIPA agreement. If they hesitate, walk away.
- Commit to disclosure. Any AI that talks to your patients should say so.
- Publish it. A single plain-language page, reviewed before launch and updated as your tools evolve.
Your patients are going to encounter AI in healthcare either way. The only question is whether they hear it from you first.
If you’re wondering how these transparency practices fit with Ontario privacy law, read our guide:
How PHIPA Applies to AI: What Every Ontario Clinic Needs to Know Before Deploying AI Tools.”
Ajax Web AI helps clinics and local businesses across Durham Region and the GTA adopt AI voice agents and automation responsibly. We automate the work, not the relationship. If you want help building your own AI transparency page and adoption toolkit, contact us.