Meta’s push to make AI-powered smart glasses an everyday technology is running into growing resistance in Australia, where regulators are calling for tougher safeguards against covert recording.
Australia’s eSafety Commissioner has urged manufacturers of camera-equipped smart glasses to automatically blur the faces of people who have not consented to being filmed and introduce clearer, tamper-resistant indicators showing when recording is taking place.
The intervention comes as Meta’s Ray-Ban smart glasses and lower-cost rivals become increasingly accessible, bringing privacy concerns around wearable cameras into sharper focus.
Regulators want privacy built in
The eSafety Commissioner argues that responsibility should not simply fall on consumers to use the devices appropriately.
Among the proposed design protections are visual recording indicators that cannot be disabled or obscured, restrictions preventing smart glasses from identifying members of the public and automatic face blurring where informed consent has not been provided.
The concerns are already more than theoretical.
Australia’s online safety regulator has highlighted reports of women being secretly recorded using smart glasses, with footage subsequently uploaded to social media and attracting sexualised or degrading commentary.
The technology nevertheless has legitimate uses, including accessibility applications for people with impaired vision, leaving regulators trying to balance those benefits against the potential for surveillance and harassment.
Meta closes recording loophole
Meta has meanwhile made its own move to address criticism of its glasses.
The company recently updated the devices so that their cameras stop operating if the front-facing recording indicator is covered while a video is being captured.
Previously, the glasses could prevent recording from starting if the LED was obstructed, but some users had discovered a workaround in which the light could be covered after recording had already begun.
Meta moved to close that loophole on August 27 and has also launched an awareness campaign highlighting the LED as the signal that the glasses are taking photographs or video.
Meta's own guidance now states that users cannot take a photo or video if the capture light is covered.
Australia weighs tougher rules
The issue could soon move beyond voluntary design changes.
Australia's federal government is preparing another round of privacy reforms that are expected to consider the rules surrounding smart glasses, although a wholesale ban is not currently expected.
The Australian Privacy Commissioner has also been examining the implications of the technology, while the Greens and some crossbenchers are reportedly pushing for a temporary 12-month ban while privacy legislation catches up.
That makes smart glasses an emerging test case for wearable AI: the same always-on cameras and sensors that make the devices useful also make them potentially far less conspicuous than pulling out a smartphone to record someone.
For Meta, which increasingly sees AI glasses as an important computing platform, convincing consumers and regulators that those cameras can be used without turning everyday eyewear into a surveillance tool may prove just as important as improving the technology itself.