Meta Ray-Ban smart glasses and privacy: what Canadian AI and wearable founders should know

Why the glasses are in the news

Meta’s Ray-Ban glasses look like ordinary sunglasses but include cameras, microphones and AI features, so the wearer can take photos and video hands-free. In August 2026, CBS News reported on a woman who found out she’d been filmed only after friends sent her the video. Meta says an LED turns on during recording, and in September 2026 it announced a change so that covering the light after recording has started also stops the camera.

For founders, the story isn’t really about Meta. It’s about what happens when a product captures people who never agreed to your terms.

Is it legal to record someone without consent in Canada?

It depends on what’s recorded and who’s doing it.

• Private conversations. Under section 184 of the Criminal Code, knowingly intercepting a private communication is an offence, but it doesn’t apply to someone who has the consent of the person who made the communication or the person it was meant for. That’s why Canada is often called a one-party consent jurisdiction.
• Public places. Recording someone in public isn’t automatically illegal, but other rules can still apply, including the Criminal Code’s voyeurism and criminal harassment provisions and civil privacy claims.
• Businesses. PIPEDA applies when an organization collects, uses or discloses personal information in the course of commercial activity. An identifiable face or voice in a video can be personal information. A company that uploads, stores or analyzes recordings generally needs meaningful consent and a clear purpose.

PIPEDA doesn’t apply to an individual collecting information purely for personal or domestic purposes (section 4(2)(b)). Creators who monetize what they film may be treated differently, so a creator running a business should get advice.

The Office of the Privacy Commissioner flagged this problem years ago. Its 2014 research paper on wearable computing gave particular attention to body-worn cameras and noted that they amplify the challenge of protecting the privacy of non-users, who may be recorded.

What happens to the data afterwards

For AI products, the recording is often only the start. Footage and audio may be uploaded, processed by models, reviewed by people to improve features, retained, or shared with vendors. Each step raises the questions users and regulators will ask: Was the person told? Where is it stored? How long is it kept? Is it used to train models?

If you have users in Quebec, its private sector privacy law requires that a technology product or service offered to the public with privacy settings have those settings set to the highest level of confidentiality by default (section 9.1), and a privacy impact assessment for projects to acquire, develop or overhaul systems handling personal information (section 3.3).

What wearable and camera-app founders can take from this

• Collect only what you need. If a feature works without storing raw video, don’t store it.
• Make capture obvious. Lights, sounds and on-screen cues help bystanders know they’re being recorded, and they should be hard to defeat.
• Explain the uses that aren’t obvious. Human review, AI training and sharing with model providers belong in plain language in your privacy policy and AI product terms.
• Set retention limits and deletion paths, including for people who aren’t your customers.
• Govern misuse. An acceptable use policy that bans covert recording and harassment gives you a basis to suspend bad actors.
• Check your vendor contracts. Data processing terms should cover every cloud and AI provider that touches recordings.

A privacy policy for a camera or wearable product should say what’s captured, from whom (including bystanders), where it’s processed and stored, whether it trains models, who it’s shared with, how long it’s kept, and how people can ask for access or deletion. See privacy and data.