Is Your Doctor Wearing AI Glasses? What to look for, What the Law Says

This Threads post about a routine cardiology follow-up (in the US, could have been Canada) drew lots of reaction. The person posting noticed partway through an EKG that the tech running the test was wearing AI-enabled glasses. The poster knew what she was looking at because she follows the gadget space. Most people don’t.

There were two broad categories of replies to this post. One group was concerned that the use of AI glasses hadn’t been disclosed - which is totally legit. The other group's concern was simpler: how do I even know someone I encounter is wearing AI glasses?

Here is some insight on both.

What are AI glasses?

AI glasses are eyewear with a built-in camera, microphone, and connection to an AI assistant. They can record photo or video, transcribing audio, and answering questions about what the wearer sees or hears. Ray-Ban Meta is currently the most common consumer version. Google, Apple, and Snap have all signaled they're building competitors, so the category is only going to get more crowded, not less.

Over time, AI glasses have evolved. In the past they were bulky, obvious gadgets. The current generation is designed specifically to pass as ordinary eyewear.

How to Spot AI glasses

Although there is no foolproof way to identify AI glasses, these three features are common:

• Frame thickness. AI glasses need to house a camera, battery, and processor inside the frame, so the temples and front rims are noticeably thicker than standard prescription or sunglasses frames.

• A small lens. Look for a small, dark, circular element embedded in the frame, usually in one upper corner. It's easy to miss unless you're looking for it.

• An LED indicator. Most current models have a small light, typically on the temple near the camera, that activates when the device is recording photo or video.

The LED is the signal people most often mention, but it is not always easy to spot. I have interacted with people wearing AI glasses – once in a bar - and never noticed the LED. What caught my attention was the pinhole lens at the top of the frame.

User behavior provides a fourth set of clues, separate from the hardware itself:

• Tapping or swiping the temple. Touch controls on these devices sit on the right temple arm. A tap, double tap, or swipe there controls music, calls, or the camera. Someone doing this repeatedly, especially with no phone in hand, is interacting with something on their face, not adjusting their glasses.

• Saying a wake phrase, then pausing. Voice-activated AI assistants on these devices respond to a spoken cue (Meta's is "Hey Meta"), confirm they're listening with an audible chime, and then the wearer goes quiet for a few seconds while the assistant responds in their ear. A one-sided, oddly clipped exchange like that, with no phone visible, is a tell.

• Holding their head still and aimed at something. The camera is fixed to the frame, not independently aimed. To photograph or record something, the wearer has to point their whole head at it and hold still for a moment, which reads differently from a normal glance.

None of these are exclusive to AI glasses on their own — anyone might tap their face or pause mid-sentence. However, these behaviors are worth noticing.

 

The risks worth knowing about AI glasses

Here's why you might be concerned that your doctor, pharmacist, store clerk, co-worker, or person you meet in a bar is wearing AI glasses. It’s not just about being recorded. It's about what happens after: where the recording goes, who can see it, and how long it sits somewhere outside your control. Here are some key risks:

• Retention and training use. Some recordings made through AI glasses or paired apps may be reviewed by humans or used to improve AI models, depending on the manufacturer's data practices and your settings.

• Audio capture, not just video. Many of these devices are listening as well as recording, even when no photo or video is being taken.

• Facial recognition. Manufacturers have explored adding facial recognition features to these devices. None have shipped that feature as of this writing, but it's a direction worth watching.

• Detection is gradually getting harder, not easier. As modification markets and stealth features grow, looking for the LED becomes a less reliable check over time.

How are we protected by Canadian Law?

This is interesting.

Canada is a one-party consent country. Under the Criminal Code, anyone who is a participant in a conversation can legally record it without telling the other person. Seems good to know.

If someone is talking to you directly while wearing AI glasses, criminal law does not require them to disclose that they're recording. What's illegal is a third party recording a conversation they're not part of.

That surprises most people, because the instinct is to assume disclosure is always required. It isn't, at least not under criminal law.

Where disclosure obligations actually originate in Canadian law is from is a separate layer: privacy law that governs organizations, not individual recording itself. PIPEDA governs how businesses collect, use, and disclose personal information. Provincial health privacy laws, including Ontario's Personal Health Information Protection Act and Alberta's Health Information Act, govern how patient information is captured in clinical settings specifically.

Ontario's Information and Privacy Commissioner issued guidance this year requiring healthcare organizations to address privacy and consent before deploying AI listening tools. British Columbia's privacy commissioner went further, requiring patient consent in nearly all clinical situations before an AI tool can be used to record an encounter.

So, the practical distinction is this: if you're at a salon, a retail counter, or a casual professional meeting, the person you're talking to generally doesn't need your permission to record you. If you're in a clinical setting, the organization has separate, more specific obligations to tell you what's being recorded and to get your consent.

The protection you have depends heavily on which kind of encounter you're in, not on whether recording without telling you feels like it should be illegal everywhere.

Questions worth asking someone wearing AI glasses

If you suspect someone you're interacting with is wearing AI glasses, these are reasonable questions to ask directly:

• Are you wearing AI glasses right now, and are they recording?

• Why are you recording and what happens to anything recorded during our interaction?

• Who else can access that recording, and how long is it kept?

• Is there a way to turn it off for this conversation?

• If this is a workplace policy, who decided that, and can I see it in writing?

Speaking from personal experience, I was very uncomfortable when I realized I was being recorded. In each case I confirmed that the person was using AI glasses. I was not comforted by them offering to turn the AI off and couldn’t verify that they had. My reaction was to shut down the conversation and move on.

Bottom Line on Encounters With AI Glasses

AI glasses aren't a novelty anymore. Your chances of encountering them are high. Just be aware they carry real risk - not because the technology is necessarily malicious - but because most of us don't know what's being captured, where it goes, or who can see it.

It's worth knowing the three tells, understanding that Canadian law generally lets the person you're talking to record without telling you, and asking the direct questions above if/when it matters to you.

It’s also worth knowing the difference between a casual encounter and a clinical one.

For business owners, leaders and Board members I also wrote this blog on the governance considerations of AI glasses and other AI recording devices that are silently making their way into all parts of business.