Big Brother on Your Nose — and Wisdom in Your Gaze

“Big Brother is watching you.” In George Orwell’s 1984, surveillance shapes people’s behaviour before those in power intervene. When people never know whether someone is watching, they begin to guard their words and actions. Smart glasses can bring a similar uncertainty to the café table, the beach and the shop. The camera sits on the nose of the person you meet. A temporary ban can be justified if it is narrow, time-limited and includes exemptions that work in practice. Lasting regulation, however, also requires practical wisdom: the ability to protect privacy while preserving technology that gives people greater independence.

Aristotle calls this wisdom phronesis. It concerns deliberating and acting well in concrete situations. Technical expertise tells us how something can be done. Phronesis helps us judge whether it should be done, why and under what circumstances. Developing glasses that can interpret a face is one thing. Understanding when that face should be left alone is another.

Smart glasses is an umbrella term for glasses with digital functions. AI glasses use artificial intelligence, for example to interpret camera images. Models equipped with cameras, microphones and AI raise particular privacy concerns. An image used to read a sign and an image used to identify a stranger may come from the same camera. Yet the actions have different moral significance.

In her Nature commentary “Smart glasses are about to create a major privacy crisis”, published on 8 October 2026, AI researcher Joy Buolamwini warns that discreet cameras are making involuntary surveillance part of everyday life. People who never chose the technology are nevertheless affected by it. When images are linked to facial recognition and databases, a face can become a gateway to a name and other personal information.

Orwell’s Big Brother was a centralised state power. Recording through smart glasses can be carried out by individuals and mediated by commercial services. The distinction matters. We do not automatically inhabit 1984 because someone buys glasses with a camera. Yet uncertainty can have a related effect: we become more guarded when we suspect that a moment is being recorded.

Privacy concerns more than secrets. It protects our ability to be vulnerable, distracted and ordinary without being turned into content. A patient in a waiting room or a child on a beach need not have anything to hide to deserve consideration. Being visible to others in the moment differs from being recorded, analysed and made accessible later.

Norwegian authorities have now announced plans to act. On 5 October 2026, the government stated that it would propose a temporary ban on the use of AI glasses in selected locations. Areas under consideration include parks, beaches, museums, shopping centres, schools, kindergartens and healthcare facilities, as well as places with changing rooms and showers. The government rejects a blanket ban and plans exemptions for vulnerable groups and socially beneficial uses. An expert group will advise on permanent regulation. This is an announcement of forthcoming legislation, not a declaration that the ban is already in force. The press release also provides no definite end date.

Existing laws already impose limits. Norwegian copyright law generally requires consent for the public reproduction or display of images of identifiable people, subject to specified exceptions. This differs from a general requirement to obtain consent before taking any photograph. Data protection rules and criminal law also apply to new technology. In August, the government stated that the EU’s AI Act had not yet been implemented in Norwegian law. It is therefore necessary to distinguish between Norwegian rules, European rules and political plans.

Data protection rules also contain an exemption for purely personal or household activities. Its scope depends on how the technology is used. The European Court of Justice’s 2014 Ryneš judgment illustrates one boundary: a fixed camera at a home that also monitored a public space fell outside the equivalent exemption in the legislation then in force. The judgment does not directly determine how brief recordings made with smart glasses should be assessed. It nevertheless shows that a private purpose does not make every recording of other people a private matter.

The EU’s AI Act also directs certain prohibitions at specific practices, including the creation of facial recognition databases through the untargeted collection of facial images from the internet or CCTV footage. Such regulation points towards an important principle: we must examine what the technology is used for and what intrusions it enables.

Here, phronesis can make the debate more precise. A blind person having a sign read aloud, a tourist filming a street and someone identifying passers-by through a database are using related technology. Their purposes, needs and consequences differ. An assessment that asks only whether the glasses have a camera is too crude. An assessment that asks only whether the user has good intentions is too weak.

Practical wisdom requires attention to the people in the situation. What does the user need help with? Who else is affected? How vulnerable are they? Could the task be accomplished by collecting less information?

This does not mean that the wise solution always lies halfway between two extremes. In a changing room, the appropriate action may be to stop using the camera. When reading a sign, it may be to permit a limited assistance function. Phronesis offers no recipe for compromise for its own sake. It requires relevant differences to matter.

For blind people and those with low vision, smart glasses can provide access to information that sighted people take for granted. They can help users read text, identify objects and understand their surroundings. The Norwegian Association of the Blind and Partially Sighted supports investigating regulation but stresses that its consequences for independence and equality must be considered. Privacy and participation are both human goods. Neither should automatically be pushed aside.

Be My Eyes makes this tension concrete. Through compatible smart glasses, a user can contact a volunteer who sees through the camera and provides spoken assistance. The user’s hands remain free. This can make an everyday task easier and reduce the need to ask people nearby for help.

But the volunteer may also see others who enter the camera’s field of view. A stranger becomes the user’s eyes while gaining a view of the surroundings. And although the volunteer’s access ends with the call, the information does not necessarily disappear. Be My Eyes’ September 2026 privacy policy describes the storage of video and audio and allows call recordings to be retained without a fixed end date. Users can request deletion, subject to certain exceptions. These are the service’s general terms, but they show that assistance in the moment can also create a lasting recording.

A good purpose does not therefore make every means good. Wise use may involve directing the camera towards the task and ending the transmission once assistance has been provided. At the same time, individuals can hardly be expected to understand data flows and further uses hidden within these services. We cannot make the user’s judgement the solution to conditions the user neither knows about nor controls.

For Aristotle, phronesis is primarily a quality of the person acting. In the debate over smart glasses, it also gives us reason to ask how institutions and products enable good actions. Clear rules and limited storage make responsible use easier. Unclear exemptions and hidden further uses make it harder.

The government’s planned exemptions must therefore work in practice. A person with low vision should not have to negotiate the use of an assistive device at every museum entrance or explain their sight loss to one employee after another. At the same time, needing assistance does not confer an unlimited right to record others in intimate situations. Exemptions and accessible alternatives must be developed together with those who need the technology.

The temporary ban should also have an explicit expiry date and be evaluated before any extension. The pause must be used to investigate misuse, test exemptions and assess whom the restrictions affect. Broad bans in parks and shopping centres require particularly careful justification. A pause that produces no learning can become a permanent restriction without a sufficiently lasting assessment of the competing interests.

Manufacturers must shoulder their share of responsibility. A visible indicator for recording and transmission is useful, but it can be overlooked in daylight or deliberately covered. It should be difficult to bypass, and the camera should stop operating if the indicator is obstructed. Even then, the light provides information, not consent.

Where the function permits it, images should be processed locally. Storage and model training should be disabled by default. Identifying information should be obscured before transmission when it is unnecessary for the task. Biometric identification of passers-by without valid consent should not be offered as a standard consumer feature.

Such requirements need enforcement. The Norwegian Data Protection Authority has a central role in matters involving personal information. The Norwegian Communications Authority, which already supervises electronic communications equipment and has been designated as the coordinating market surveillance authority for the AI Act, is a natural candidate to oversee relevant technical requirements. New duties nevertheless need a clear legal mandate. Responsibility must be assignable.

Phronesis supplies no ready-made regulation. It offers a standard of judgement: do our actions and rules promote a good life together? Smart glasses should be assessed according to their effects on independence, dignity and trust. Those assessments must also remain open to revision as experience becomes clearer.

“Big Brother is watching you” reminds us of what is at stake when people lose the space to go unnoticed. The blind user reminds us of what is at stake when people lack access to the world. Practical wisdom requires us to keep both experiences in view. We can then protect ourselves against unwanted scrutiny while preserving the assistance that makes participation possible.

Starting point: Joy Buolamwini, “Smart glasses are about to create a major privacy crisis”, Nature, volume 658, page 296, 8 October 2026. https://doi.org/10.1038/d41586-026-03131-x.