If you’ve been following me for a while, you know I talk a lot about reviewing your visit summaries. Read them. Check them for accuracy. Flag anything that doesn’t look right and reach out to your provider to correct it. This has always been important because your medical records follow you for life. And a new report from Ontario’s Auditor General just made the case more critical than ever.
Have you ever been asked by a provider if they can use an AI scribe tool?
In May 2026, Auditor General Shelley Spence released a special report on the use of artificial intelligence across Ontario’s public service, including a deep look at AI scribe programs being used in healthcare settings. AI scribes listen to your appointment and generate clinical notes on your doctor’s behalf. They’re being adopted quickly across North America as a way to reduce administrative burden on providers, which sounds great… in theory.
Here’s what the audit found:
Out of 20 AI scribe systems evaluated, nine fabricated information entirely. We’re talkin’ full-blown nonsense where the AI-generated referrals, test orders, and treatment suggestions were listed but never discussed during the appointment. Twelve of those same 20 systems recorded a different drug than the one the doctor actually prescribed. Yikes! Seventeen missed key details about patients’ mental health that were shared during the visit.
And these were government-approved systems. Yet more than half of them captured the wrong medication.
Here is the part that made my eyeballs nearly fall out of my skull. Doctors in Ontario were not required to officially sign off on the AI-generated notes to confirm their accuracy. That means a note could go into your chart, or be used to inform your care, without your provider ever formally checking that it was correct.
One thing I need to call out. This audit examined models from 2024, and the technology is evolving. Some in the healthcare AI space have pointed out that current versions may perform differently. But here’s the reality for most of us: we have no way of knowing which version our clinic is using, whether it has been updated, or whether our provider reviewed the output before it became part of our record.
And that is exactly why you have to be in the loop ALWAYS.
Reviewing your visit summary has always been a best practice. Now it’s a non-negotiable.
After every appointment, open your patient portal and read your summary. It can take hours to weeks for the visit summary to show up.
Here’s your review checklist:
Does the medication list reflect what was actually discussed? If your doctor changed a dose, added something, or discontinued a medication, does the note say so correctly?
Does the assessment reflect what you actually talked about? AI scribes generate structured notes. If something was said in the room but isn’t in the note, or if something appears in the note that was never discussed, that is a problem worth addressing.
Does the plan make sense? Referrals, follow-ups, lab orders. Do they match your understanding of what comes next?
If something is wrong, contact your provider’s office. You do not need to be aggressive or accusatory. A simple message through your portal works: “I reviewed my visit summary and wanted to flag something that in accurate/doesn’t reflect our conversation and I would appreciate it updated. Thank you.” That’s it. Most offices want to know.
This is one of the most direct ways you can protect yourself in a healthcare system that is moving very fast. Technology is being adopted into clinical settings at a pace that sometimes outstrips the guardrails. Patients often end up being the last line of verification. I know it shouldn’t work that way, but that is the way it works.
You were in that room. You know what was said. Trust that.
The information shared in this post reflects my experience as a board-certified patient advocate. It is not intended as medical advice and should not replace guidance from your own healthcare team.
Little Engine Patient Advocacy 
Leave a comment