Meta is reportedly developing a new prototype of smart glasses that incorporates artificial intelligence (AI) functionalities and has the ability to constantly monitor the environment.
'Super Sensing' Project Features
As reported by the Financial Times, this internal project, named 'super sensing,' envisions devices capable of continuously recording audio and taking photographs at short intervals. This functionality would allow users to later consult Meta AI about recorded events, using the images and sounds collected throughout the day as a basis for answering their questions.
Metadata Alternative for Privacy
One of the architectures discussed by Meta suggests that raw audio and video files should not be stored by the company itself nor made directly available to the user. In this scenario, only the metadata extracted from this content would be sent to the company's servers, where it would be processed by Meta AI to meet the user's requests. Proponents of this methodology argue that it presents lower privacy risks compared to storing complete recordings.
Privacy Concerns and Issues
If the glasses or the 'super sensing' features are commercially launched, it is expected to intensify the public debate on privacy. Meta already faces scrutiny due to its work on facial recognition technologies for its smart glasses. Furthermore, the company received criticism after reports of users filming women while wearing the device and dealt with modifications made by third parties that removed the recording indicator LED.
In response to these issues, the company announced an update on Tuesday (7) that automatically disables the camera if it detects that the light indicator has been altered. However, the Financial Times points out that Meta's current plan foresees the LED remaining off during the operation of 'super sensing' mode.
Light Indicator Strategy
A technical document from Meta, published in July 2025, already detailed this strategy. According to the white paper, the light indicator would be reserved exclusively for moments of 'active capture,' meaning when the user decides to save photos or videos. During the use of features classified as 'AI Feature,' such as scanning a menu with AI, the LED would remain inactive. Meta justifies this measure by saying it prevents people from getting used to ignoring the indicator.
The Financial Times also notes that if the LED were lit up during 'super sensing' mode, it would be more complicated to discern whether the glasses were merely analyzing the environment or actually recording videos. Additionally, Meta is internally studying the possibility of using data captured by the glasses to train its AI models. The newspaper also reports that the 'super sensing' features can be implemented in the company's existing smart glass models.
Meta's Statements on Privacy
Dave Arnold, a Meta spokesperson, avoided commenting directly on the project in a statement sent to The Verge. He stated: 'While we do not comment on internal prototypes, we are committed to getting our glasses right because they need to be loved by both the people who wear them and those around them.'
Arnold added that 'our approach has been to develop new technologies that help people throughout the day, with privacy built in from the start.'Future Vision of Personal Assistants
Even without confirming the new prototype, Meta had already expressed interest in smart glasses with continuous operation. In a first-quarter 2026 earnings call, CEO Mark Zuckerberg expressed enthusiasm about seeing glasses evolve from mere question responders to become an all-day personal agent, assisting with memory and goal achievement.
In a March publication about the Ray-Ban Meta glasses, the company also announced that, through software updates, Meta AI will become 'a more continuous assistant, present in the moment, capable of helping throughout the day,' overcoming the need for explicit user commands.
