Meta is reportedly testing ‘super sensing’ AI glasses that could remember far more of what the wearer sees and hears. Camera glasses are also beginning to find workplace uses, where persistent sensing would raise harder questions around privacy, company data and who controls when the devices are listening or looking.
Meta is reportedly testing AI glasses designed to remember far more about the wearer’s day.
The Financial Times reports that prototype systems could continuously collect audio and take photographs every few seconds, giving Meta AI more context about what a user has seen, heard or done. The report says Meta executives are considering a system where the capture LED would remain off during the background sensing feature, although the plans could still change.
Meta hasn’t announced the reported “super sensing” system as a product. The FT also reports that the capabilities could potentially reach existing Meta glasses through software updates, although no such rollout has been confirmed.
Privacy Concerns Already Affect How People Behave
Speaking to The XR Beat before the FT report was published, ArborXR Enterprise Growth Lead Devin Marble described being asked to remove his Meta glasses during a discussion at AWE.
“I’ve been in conversations with people at AWE where I was wearing my Meta Ray-Ban displays, and they said, ‘Could you take those off?’ because they didn’t trust that they weren’t being recorded or photographed while we were having a discussion.”
Marble said the concern remained despite the presence of a capture light.
Meta published a privacy FAQ for its AI glasses this week, saying its current products use a white capture LED when content is being captured for the wearer’s gallery. Covering the LED or physically tampering with it disables the camera, according to the company.
Meta says wearable technology needs the trust of the wearer and the people around them. Its current capture LED is designed to give bystanders a visible indication that a photo or video is being taken.
The reported super-sensing concept raises a harder issue because its purpose is different from deliberately taking a photograph or starting a video recording.
According to the FT, one proposed system would extract information from audio and images for the AI to query without storing the raw footage and audio for the user. Meta is also discussing whether information collected through the glasses could be used for AI training.
People around the wearer may care about the information being gathered even if the system handles it differently from conventional video recording.
Consumer Glasses Are Already Finding Professional Uses
Meta’s current AI glasses are consumer devices, and Meta itself describes the range as a fast-growing consumer product used for music, photos, video and AI assistance.
But some of the same hardware is already being adapted for professional workflows.
ArborXR recently became the first MDM platform to support Meta AI glasses through its Remote Assistance product. A remote expert can see a live stream from the wearer’s camera and communicate through two-way audio.
Remote assistance is already one of the more established enterprise AR workflows, with platforms such as TeamViewer developing similar hands-free support models.
The use case is deliberately narrow; a worker starts a remote support session for a specific job and another person sees what the wearer sees.
In The XR Beat interview, ArborXR Director of Channel Partnerships EMEA Toriâh Teague described meeting buyers who were spending heavily to fly specialists between factories to help solve problems and train workers.
One example involved specialists regularly travelling between Montreal and the UK to support work on site. Teague said companies were looking at remote assistance as a way to provide expert support without putting someone on a plane each time a problem arose.
ArborXR’s current Remote Assistance support has no connection to Meta’s reported super-sensing prototypes. It shows how hardware designed for consumers can move into professional guided workflow use when the camera, audio and form factor suit a specific job.
Workplace Deployments Need More Than a Capture Light
A remote assistance session has a clear start and end. The worker turns it on for a specific task.
Background sensing could work very differently. A device worn for much of the day can encounter colleagues, customer information, computer screens, documents, private discussions and proprietary equipment.
Companies need to know when cameras and microphones are active, what information is processed and who can access it.
Marble told The XR Beat that smart glasses deployments need controls around identity, location, company data and what happens when a device leaves an approved environment.
“We need to help organisations wrap their heads around what is connected to these glasses. When are they on? When are they off? What happens if they walk off site?”
He also discussed authentication, geofencing and the ability to shut down, reboot or wipe devices remotely.
Comfort can complicate those controls.
Marble described the possibility of a worker forgetting they were wearing the glasses and carrying them into an inappropriate area. A conventional headset is generally put on for a deliberate session. Lightweight glasses can stay with someone as they move between different parts of a site.
AI creates further concerns around company information. Marble said organisations need to consider which AI systems can access internal data and where processing happens when glasses are seeing company equipment, documents or other intellectual property.
Authentication, permissions, data control and remote management are familiar requirements for phones and laptops, and already part of the XR enterprise stack. Smart glasses combine some of those requirements with cameras and microphones worn directly on the face.
The reported Meta system remains a consumer-focused prototype. No enterprise version has been announced, but the FT reports that the technology could be deployed to existing glasses through software updates.
If deeper background sensing reaches devices that later find professional uses, businesses may need the ability to restrict individual features rather than simply deciding whether a pair of glasses is allowed on site.
Different Glasses, Different Privacy Trade-Offs
Meta’s reported work arrives as smart glasses companies take very different approaches to cameras and sensing.
Even Realities recently raised $150 million at a $1 billion valuation for a camera-free, display-first model. Its G2 glasses show information including translation, navigation, teleprompting and AI assistance through an in-view display.
Meta is pursuing hardware built around cameras, microphones and AI that can understand more about the wearer’s surroundings.
The two approaches create different privacy and workplace considerations.
Camera-led glasses can support remote assistance, visual inspection and AI applications that need to understand physical objects. Companies using those functions need controls over where cameras operate, what information is processed and when workers or other people nearby need to be informed.
Camera-free glasses avoid some camera-related concerns, while microphones, transcription and AI processing still need clear policies and consent.
Meta’s reported prototypes remain under development, and no super-sensing product has been announced.
The immediate privacy debate is about Meta’s ambitions for consumer AI glasses. Workplace teams have a separate issue developing alongside it: consumer hardware can now pick up professional use cases quickly, and companies may need more control over which capabilities remain active once the glasses come through the door.
Images from Meta
