AI Glasses Hit A Privacy Wall Before They Reach Everyone

The glasses are arriving before public trust. Courts, schools, testing centers, cinemas and other public venues are building rules around AI eyewear, turning privacy concerns into a growing patchwork of bans.

No country has imposed a permanent, outright ban on AI glasses. Pakistan’s Punjab province came close with a 30-day order that began in October 2026, covering camera-equipped smart glasses with “hidden recording ability” as well as miniature cameras disguised as pens or keychains.

The Punjab order covers sales as well as use, including online sales. Punjab authorities said they received complaints about spying devices being used for illegal surveillance, harassment and blackmail. The restriction applies only to Punjab, but the message is not subtle: a wearable camera can become a public-order problem before it becomes a mass-market success.

Where smart glasses are already restricted

New York banned smart glasses with cameras, microphones or computers from all 1,240 state and local courts on July 20, 2026. Visitors must turn them over to uniformed court officers for safekeeping, and switching them off does not solve the problem. Court officials are urging visitors to bring a traditional pair instead.

The policy followed a February 2026 appearance in Los Angeles, where Mark Zuckerberg’s team wore Ray-Ban Meta glasses during his testimony. Judge Carolyn Kuhl warned, “If your glasses are recording, you must take them off,” although no recording of the proceedings was confirmed. Meta glasses face similar restrictions in courts and tribunals across England and Wales, where security personnel can hold them until visitors leave.

Schools are another obvious flashpoint. Western Australia banned AI glasses from public schools on October 12, 2026, covering students, staff and visitors while allowing exemptions for approved medical, accessibility and educational reasons. The State of Victoria announced similar restrictions, set to take effect on January 27, 2027.

Smart glasses are also prohibited at SAT test sites, including during breaks unless an accommodation is granted, and scores may be canceled for violations. Marine Corps policy bars their use at Camp Lejeune and other bases, while Canterbury-Bankstown Council has banned them at leisure and aquatic centers.

UK cinemas are rolling out measures against camera-enabled glasses because of privacy and film piracy concerns. Royal Caribbean permits them on board but prohibits their use in casinos, medical facilities and youth areas. Norway’s government is considering temporary restrictions that could cover parks, beaches, shopping centers and gyms, though those proposals have not been enacted and officials have not settled which glasses would qualify.

Paris prosecutors opened investigations in September 2026 after complaints that women were filmed without consent using smart glasses, with the footage posted online. Prosecutors did not identify the brands involved. That distinction matters less to bystanders than it does to product lawyers: if the camera is invisible in practice, the brand on the frame will not rescue public confidence.

Google wants privacy to be part of the product

Google launches its intelligent eyewear devices today after working on the platform for years. Its privacy plan reaches beyond a light on the frame, because a blinking indicator alone is not a policy — it is a tiny apology waiting to fail.

The glasses use an LED to show when someone is taking a picture or recording video. Sensors detect whether the light is covered or tampered with, and the glasses prevent image or video capture when that happens. Juston Payne, Google’s lead for XR, said, “Unfortunately, some of the glasses that have launched in the past launched without any kind of detection for those things, and that broke trust. And that’s a problem.”

Google says camera visuals are not stored in Gemini history during Gemini processing, and the glasses are designed to automatically redact bystanders’ faces before Gemini processes those visuals. The camera stream will not be used for AI training or ad targeting, while data sent to the cloud or a phone for processing is encrypted in transit.

Google also plans to bring its Private AI Compute platform to the glasses, using a “hardware-secured sealed cloud environment” for processing. Four microphones use beamforming and directional focus to identify the person being addressed. Google says Gemini is designed to focus on the wearer’s voice rather than nearby voices, while Payne said the microphones filter out “all the world audio, as we call it, the bystander audio.”

Users can connect the eyewear to Messages, Keep and YouTube Music to send replies or create lists, but each app must receive permission before accessing the camera. Users can add or revoke access for every app, review past Gemini interactions, delete individual chats or all chats, and decide whether conversations are used to train AI.

The controls continue outside the device. Google Shopping and Google Ads will ban the sale and advertising of stickers or accessories designed to hide recording lights. Play Store and API policies will prohibit apps from using the camera for social livestreaming, real-time facial identification or changes to the outer indicator light. YouTube will prohibit content depicting harassment and unwanted sexualization, regardless of whether the footage came from a phone, camera or glasses.

Google requires an 18+ Google account to activate the eyewear and asks users to agree not to record others without permission or in sensitive places. Payne said there are “multiple approval layers if an app wants to use the camera,” along with use cases Google is “just blanket cutting off.” He also admitted that “there will always be ways around what we build, and then we’ll try to fix those holes.”

That ongoing back-and-forth will decide whether the category survives. Meta claims its white recording LED lights during photos and videos, but Ray-Ban documentation says some camera-based AI functions on newer models can work without the LED because they do not save shareable photos or videos. That is not the same as stealth video, but bystanders may not know when the camera is being used for AI.

The practical case for smart glasses is still unsettled, too. They can provide privacy from people peering over a shoulder, declutter a workspace and create multiple virtual monitors. Yet testers have reported eye strain, display limits, heat, awkward video calls, isolation at work and difficulty switching tasks or writing in a physical notebook.

Stephen Johnson said his Xreal One Pro glasses heated up during long sessions, while Jasmine Uniza found hers comfortable for about an hour and a half to two hours. Cheaper glasses may lack the resolution needed for text or precise work, and several testers found the devices better for gaming and movies than office work.

Smart glasses have become lighter and more streamlined. For now, though, they look better suited to movie night than the workday — and public trust remains the monitor they cannot replace.

Stay connected via Google News

Follow us for the latest AI News