We suppose it was inevitable, but it didn’t take long for people to figure out how to turn smart glasses into a privacy nightmare.
What happened: Two Harvard students rigged a pair of second-generation Ray-Ban Meta smart glasses with facial recognition software and some online databases to rapidly identify strangers in public.
- The students sent a feed from the glasses to a tool that finds people’s images online. From those images, they then used AI to determine the person’s name and compile publicly available information about them in an app they wrote.
- In a demonstration video posted on X (which, trust us, is worth watching), the students used the glasses to identify people on public transit and get detailed info about their personal lives, including their job and volunteer roles.
Why it matters: It’s a powerful example of how the convergence of several technological advancements — inconspicuous and always-on cameras, AI-powered data scraping, and facial recognition software — poses a serious risk to people’s privacy.
- Googling someone you just met is nothing new, but being able to gather so much information about total strangers almost instantly is uncharted — and potentially risky — territory.
What’s next: The students are not releasing their code, saying the experiment was meant to call attention to the risks of the technology. Some companies, however, are pushing forward.
- Clearview AI, which makes facial recognition software for law enforcement, plans to release smart glasses with built-in facial recognition.
Hot tip: This document explains how to remove yourself from the search engines used to identify people in the experiment.—TS