The ‘smart glasses’ label now covers camera assistants, display eyewear, audio wearables and industrial systems with different jobs and risks. Increasing privacy pressure could force the industry to name them properly.
We may soon look back on “smart glasses” as a temporary name for several products that had little in common beyond their frames.
Current smart glasses privacy disputes expose how loose the label has become. A Ray-Ban Meta model with cameras and visual AI carries it, but so does camera-free display eyewear built for prompts and translation. Audio glasses for calls and music, and rugged devices used for remote assistance on factory floors are all placed in the same category.
The hardware, permissions and jobs are already diverging. Europe’s scrutiny of Meta’s camera glasses and a new policy model from Kobe University may force the language to catch up.
Smart Glasses Privacy Rules Are Targeting Specific Hardware
Hamburg’s data protection commissioner, Thomas Fuchs, says Meta’s recording light isn’t visible enough in normal use. His authority tested the glasses and concluded that recording people in public can breach data protection law. Fuchs believes a full German ban could be possible, although no ban has been announced.
France’s CNIL surveyed 2,128 adults at the start of 2026 and found that 67% considered smart glasses a privacy risk. The regulator warned that people nearby may struggle to know whether ordinary-looking frames are capturing their image, voice or behaviour.
Camera-equipped glasses create this specific risk because a wearer can record without taking out and aiming a separate device. Camera-free display eyewear avoids covert photography, although microphones, transcription and cloud AI introduce separate consent and data questions.
New York’s court system has already adopted a hardware-based rule. Its restriction covers recording-capable glasses at the door. A display-only pair and a camera pair may look similar, but only one falls within the policy.
One Label. Four Product Types
Meta’s success proves camera glasses have a large consumer market. The company held 69.2% of the smart glasses market in the first quarter of 2026, according to IDC figures previously covered by The XR Beat.
The category becomes less well-defined once products are compared by hardware and use. Kobe University’s newly published six-layer, 21-measure privacy model includes separate checklists for four types:
- Audio glasses
- Camera glasses without displays
- Camera-free display glasses
- Glasses containing both cameras and displays
Even a privacy framework has to split “smart glasses” into more precise groups before it can recommend suitable controls.
Camera glasses or visual AI glasses can centre on photography, video and assistants that interpret the wearer’s surroundings.
Display glasses can supply prompts, navigation, translation and notifications without visual capture. Even Realities has made camera-free design part of its identity while raising $150 million around its display-first model.
Audio glasses can compete with earbuds and assistants through calls, music, translation and voice AI. Their privacy risks come mainly from microphones, processing and retention.
Industrial AR devices can retain cameras because remote experts, inspectors and field teams need to see equipment or record evidence. Buyers judge these products through workflow fit, management controls and approved locations rather than consumer photography.
The final names may differ, but “smart” says little about which of these products someone is wearing or what it can do.
Smart Glasses Privacy Could Create Further Categories
Kobe University’s proposal goes beyond simple recording lights. The team recommends omitting dangerous software functions such as facial recognition, automatically restricting capture and sharing according to location or situation, masking bystanders’ faces in real time, accepting electronic refusal signals and subjecting high-risk uses to independent audits.
Two camera-equipped pairs could therefore serve very different markets. One may leave recording decisions to the wearer, whereas another could disable capture in a hospital, and perhaps respond to a no-capture beacon in a secure facility and keep an auditable record of how it was used.
Camera-free products can also separate. Halliday’s camera-free G2 uses a four-microphone array and Meeting Flow features for live transcription and summaries. A device can remove visual recording while still processing sensitive speech.
Smart glasses privacy checks need clear information about sensors, software permissions and administrative controls. Buyers need to know whether they’re evaluating capture eyewear, display eyewear, an audio assistant or managed work equipment before privacy checks can become meaningful.
Buyers Need Well-Defined Labels and Use Cases
“Smart glasses” may survive as an umbrella term, in the same way that “wearables” does, but product pages, workplace policies and procurement briefs will need more focused language.
A hospital may want camera-free prompts with tightly controlled transcription. A manufacturer may need a managed camera for remote assistance. A consumer may buy visual AI glasses because photography is the main attraction. Each buyer is choosing a different combination of sensors, software and permissions.
The products are multiplying while the catch-all name becomes less precise. A hospital choosing camera-free prompts and a factory buying a camera for remote assistance should no longer be treated as buyers of the same product merely because both devices sit on the nose.
