Restaurants, pubs and theatres have also introduced their own restrictions. These policies reflect a practical problem: staff may not be able to tell whether a customer is simply wearing glasses or recording the people around them.
The UK response shows why the technology is difficult to regulate through ordinary camera rules. A venue can prohibit filming, but it may also decide that identifying whether filming is happening is too difficult and restrict the device itself.
On August 18, 2026, Immigration and Customs Enforcement reportedly barred employees from wearing Meta glasses or similar devices in federal workplaces. The agency treated the glasses as personally owned body-worn cameras and warned that they could unintentionally capture, record or transmit sensitive information, potentially compromising privacy and legal protections.
The policy followed reports that ICE and Customs and Border Protection personnel had been seen wearing Ray-Ban Meta glasses during immigration operations in several states.
The US situation also exposes a policy tension. DHS’s proposed fiscal 2027 budget includes $7.5 million for operational prototypes of smart glasses with real-time biometric-identification capabilities for ICE field agents. The project remains at the development and budget-proposal stage, but its stated purpose distinguishes agency-controlled equipment from consumer devices used without an official recording system or chain of custody.
That distinction addresses workplace control, not every civil-liberties concern. Government-issued facial-recognition glasses could raise their own questions about accuracy, oversight, data retention and the rights of people identified in the field. The supplied sources establish the proposed capability and funding, but do not establish that such a system is already deployed.
In Germany, digital-rights organization HateAid filed a criminal complaint on August 12 against Meta, Ray-Ban and Oakley entities and several retailers selling the glasses. HateAid argues that smart glasses resembling ordinary eyewear while enabling unnoticed recording may conflict with German restrictions on devices disguised as everyday objects and intended to record people covertly.
HateAid has called for a sales ban on the Wayfarer Gen 2 model and for binding “safety by design” requirements intended to prevent people from being filmed without their knowledge.
A criminal complaint is an allegation and request for investigation—not a conviction, a final judicial ruling or proof that sales have already been prohibited. The legal significance of the German action is that it challenges the sale and design of the product, rather than focusing only on how an individual wearer behaves.
Meta has disputed the premise, saying the glasses were approved by Germany’s regulator in 2022 and were designed with privacy safeguards. The competing positions illustrate the core legal question: whether an obvious recording indicator and other safeguards are sufficient when the camera itself is built into an object that looks like normal eyewear.
The debate intensified after reporting about facial-recognition code associated with Meta’s glasses app and a surfaced patent describing smart cameras that could identify people or interpret scenes. A patent shows a technology company’s contemplated design, not proof that the patented feature is active in consumer products.
The Electronic Frontier Foundation said it confirmed the presence of facial-recognition code through static analysis and later reported that Meta removed unactivated facial-recognition code from a subsequent app update after public scrutiny.
Even a dormant or removed feature matters to critics because it changes the potential use of the camera. Recording creates a visual archive; automated identification could connect that archive to names, profiles or other personal information. That is why digital-rights advocates describe the issue as one of surveillance and practical anonymity, not only covert photography.
The controversy is unfolding as Meta’s smart-glasses business expands rapidly. EssilorLuxottica reported more than 7 million AI glasses sold in 2025, compared with about 2 million sold across 2023 and 2024 combined.
At that scale, even a small share of abusive or careless use can affect large numbers of people. More devices also make camera-equipped eyewear harder to treat as an unusual exception. The public question becomes whether people should have to assume that ordinary-looking glasses may be recording them in cinemas, restaurants, workplaces and on the street.
The UK, US and German responses differ, but they point to the same design problem. Institutions are trying to preserve spaces where recording is restricted, confidential information is protected and people can move through public life without being individually identified. Consumer smart glasses make those expectations harder to enforce because the camera is wearable, mobile and socially ambiguous.
The immediate measures are therefore practical rather than comprehensive: restrict the glasses at the door, treat them as recording devices at work, or challenge their sale under laws aimed at covert equipment. Whether those measures are enough will depend on safeguards that make recording genuinely visible, limit downstream identification and give bystanders meaningful protection—not simply on whether the wearer intends to behave responsibly.