Meta AI Glasses Are Harder to Avoid. Detection Apps Are No Cure

Introduction

Camera-equipped glasses are moving from novelty to a familiar consumer product. That shift changes the privacy question for everyone nearby: not whether the wearer is allowed to record, but whether other people can know that recording might be happening. Ars Technica’s report examines the growing backlash around Meta AI glasses and the hobbyist tools being built to detect them.

Key points

• More venues are imposing restrictions. Schools, courts, restaurants, and entertainment venues have begun banning or limiting smart glasses. DEF CON 2026 was reportedly among the events that adopted a no-exceptions ban, telling attendees who needed prescription lenses to bring non-smart eyewear.
• The built-in warning is easy to defeat. Meta glasses use a small recording light. The Electronic Frontier Foundation’s Cooper Quintin told Ars that a light that can be covered with a sticker is unlikely to deter someone determined to misuse the device.
• Detection apps provide only an indication. Nearby Glasses, Antizuck Smart Glasses Scanner, and the newer free app Zuckoff look for Bluetooth signals associated with glasses paired to a phone. They cannot establish that a camera or microphone is active. Walls, people, interference, and false positives can also make scans unreliable.
• Future features could raise the stakes. EFF has warned that facial recognition and more continuous sensing could create additional biometric and civil-rights risks, including harassment, deepfake abuse, or nonconsensual intimate imagery. Meta also removed an unreleased facial-recognition system from its smart-glasses app after it was reported, although its future plans remain unclear.

Why it matters

The popularity of detector apps reflects a basic social demand: people want some way to respond when recording may be taking place without their consent. The developers generally make a modest promise. A scan may help someone decide whether to enter a sensitive space, or help a venue check whether its smart-glasses policy is being followed. It is not a guarantee of privacy.

That distinction is crucial. A Bluetooth alert usually means only that a compatible device may be nearby and connected. It says nothing about whether the wearer is recording, whether the signal belongs to the glasses in question, or whether the scan missed a device behind an obstruction. App makers therefore warn users not to confront people based on a result and not to treat the software as a security system.

The problem may become harder as smart glasses adopt less conspicuous designs. Today, many models can still be recognized by sight, but that advantage may disappear as manufacturers offer more fashionable frames. Reports about a possible “super-sensing” prototype that could capture a wearer’s experiences more continuously have intensified concerns, particularly because sources claimed its recording light might not be activated for every feature.

Detection tools can still play a useful supporting role, but the burden cannot be placed on bystanders. Durable protection requires clear venue rules, hardware indicators that cannot be easily bypassed, responsible handling of uploaded video, and laws that give people meaningful remedies for biometric and privacy violations. Apps can make an invisible risk more visible; they cannot replace consent, accountability, or regulation.

Source: Ars Technica AI