Meta has begun selling its artificial intelligence (AI) glasses in Japan, raising mounting concerns over privacy violations and illicit recording. The product's design, nearly indistinguishable from ordinary eyewear, makes it difficult for bystanders to notice when they are being filmed.
According to Japan's Nihon Keizai Shimbun (Nikkei) on Wednesday, Meta recently launched the AI glasses "Ray-Ban Meta" in Japan. Developed jointly with French eyewear maker EssilorLuxottica, the device features a camera and speakers embedded in the frame.
Users can take photos through voice commands alone, search for objects within their field of view, or translate text, among other AI functions.
The concern lies in the device's appearance. Because it looks little different from ordinary glasses—unlike pulling out a smartphone to take a photo—people nearby find it hard to notice when they are being filmed. Critics warn the device could be exploited for illicit recording, such as covert photography. Corporate security has also come under scrutiny, as wearers in meeting rooms or workplaces could photograph documents or monitor screens, potentially leaking internal company materials.
Experts believe the potential for misuse will grow further if AI glasses are combined with earphones or lens-type displays. Users could show exam questions or game situations such as Go, shogi, or chess to the AI in real time and receive advice—a clear avenue for cheating.
Conscious of privacy backlash, Meta designed the glasses so that a light-emitting diode (LED) on the front activates during recording. The camera is also programmed to stop functioning if the LED is covered.
Yet doubts remain over whether these safeguards are sufficient. In the United States, tape designed to enable recording even when the LED indicator is covered is already being sold. Services that modify the glasses to prevent the LED from lighting up have also reportedly emerged—a cycle of technical safeguards and attempts to circumvent them.
Competition in the AI glasses market is also intensifying. Google is preparing to launch its own glasses-type AI device this autumn. Reflecting on the experience of "Google Glass," which was effectively pushed out of the market amid privacy controversies, the company said it is developing the new product with thorough consideration of the privacy of both users and bystanders.
The greatest concern is the potential future integration of facial recognition. Ray-Ban Meta currently does not include facial recognition, but if AI glasses are combined with such technology, users could instantly identify strangers they encounter on the street.
Those who have publicly disclosed face photos and personal information on social networking services (SNS) could be especially vulnerable. By cross-referencing faces captured through AI glasses with external facial recognition systems and publicly available online information, users could potentially identify a person's name, alma mater, address, phone number, and even family relationships.
In 2024, two students in the United States conducted an experiment combining footage filmed with AI glasses and an external facial recognition system. The experiment reportedly identified some subjects' personal identities and lifestyle information through publicly available online data.
In response, more than 70 civic groups including the American Civil Liberties Union (ACLU) have called on Meta to refrain from introducing facial recognition functions, warning that the technology could seriously infringe on personal privacy.
Experts argue that given the rapid spread of AI glasses, self-regulation by individual companies alone has clear limits. "An industry-wide common indicator system needs to be established so that people nearby can recognize not only when recording is taking place but also when AI functions are in use," said Tsutomu Terada, a professor at Kobe University.
