Feed: Engadget is a web magazine with obsessive daily coverage of everything new in gadgets and consumer electronics
Feed Link: https://www.engadget.com/
---

Title: Meta is reportedly working on facial recognition for its AI glasses

Date: Wed, 07 May 2025 19:55:02 +0000
Link: https://www.engadget.com/wearables/meta-is-re...

Diminished tech privacy appears to be another ripple effect from Trump 2.0.
The Information reported on Wednesday that Meta has changed its tune on
facial recognition. After considering but ultimately bailing on the
technology for the first version of its smart glasses, the company is now
actively working on wearables that can identify nearby faces. Remember when
being a "Glasshole" was considered a faux pas?

According to The Information, Meta has recently discussed adding software to
its smart glasses that scans bystanders' faces and identifies people by name.
The company has also reportedly considered adding the tech to future AI-
powered earphones with built-in cameras.

Facial recognition tech would be part of a Meta feature dubbed internally as
"super sensing." The feature would build on the glasses' live AI feature,
which can only remain active for around half an hour (thanks to its battery).
But in future devices, expected in 2026, it could run for hours.

Sam Rutherford for Engadget

The Information says the super-sensing mode wouldn't likely be the default
mode for the glasses. The glasses owner would have to opt in. But the
person's nearby company ΓÇö those having their faces scanned and named ΓÇö
wouldn't.

Making matters worse, bystanders might not even know they're being scanned.
The current Ray-Ban Meta glasses turn on a light while recording. It's a
privacy-focused feature likely spawned from Big Tech's lessons from the
Google Glass social backlash.

But Meta is reportedly questioning whether future glasses should activate the
light when the device is "super-sensing" them. Uh oh.

Putting on a pair of glasses that gives you AI-powered superhuman memory
might sound pretty cool. No need to remember things ΓÇö just let the AI scan
your environment and remind you! But the tech sounds a lot less fun when you
think about the poor souls in proximity to one of these, uh, Metaholes.

Karissa Bell for Engadget

Alongside the revival of facial recognition, Meta has updated its privacy
policies. In April, the company changed its terms so that its current smart
glasses will activate AI by default. The only way to opt out is to deactivate
the "Hey Meta!" trigger phrase. Adding to the fun is a change that no longer
lets glasses owners opt out of allowing the company to store and train on
their voice recordings.

The Information draws a line from Trump's reelection to Meta's ethically
dubious changes. The current Federal Trade Commission (FTC) isn't keen on
regulations that slow big businesses' profits. Last month, FTC Commissioner
Melissa Holyoak promised a "flexible, risk-based approach to privacy
enforcement." The agency has also stopped using labels like "surveillance
advertising."

This article originally appeared on Engadget at
https://www.engadget.com/wearables/meta-is-re...
recognition-for-its-ai-glasses-195502788.html?src=rss

---
VRSS v2.1.180528