In case you haven’t seen it, there’s been a lot of media over the last few weeks about the “privacy nightmare” of smart glasses. Cheap camera glasses recently sold out at Kmart, stories followed about people being photographed without their knowledge, and privacy advocates began warning about the consequences of putting increasingly discreet recording technology on people’s faces.
The concern is real. A camera hidden inside something worn on your face creates obvious opportunities for people to behave appallingly, and I have no particular desire to live in a world where bathrooms, bedrooms and private conversations become fair game because someone found a sufficiently small lens with a battery.
I do, however, find much of the reporting frustratingly simplistic. I’d like to take a step back from the more alarmist coverage and ask: what, exactly, are we worried about?
I’m blind. More accurately, I describe myself as blind with just enough vision to be dangerous. That means I already move through the world without reliably knowing whether a camera is pointed at me. I won’t notice the security camera mounted in the corner, and I won’t see someone across the bus lift their phone, turn it in my direction and press record.
I’m pretty sure most of the time the recording is aimed at my beautiful guide dog, and not me. But it’s still something I don’t usually notice.
One of the things people seem to find particularly unsettling about smart glasses is that they remove those familiar visual cues. Perhaps that changes the way the issue feels to me, because I mostly can’t see the cue anyway.
It doesn’t mean I care less about privacy.
It does make me wonder whether we have been relying on the visibility of cameras as a rather flimsy substitute for privacy all along.
What smart glasses actually change
Smart glasses clearly change something, but the ability to record another person without their consent is hardly new.
Phones have cameras. Laptops have cameras. Doorbells have cameras. Cars have cameras. Shopping centres have cameras. Tiny covert cameras have been sold inside pens, buttons, clocks, keyrings, teddy bears and objects with increasingly imaginative reasons for containing a camera, for decades.
Humanity did not arrive at 2026 blissfully unable to record another person until the camera moved from their hand to their face.
What has changed is the friction around doing it.
A phone usually requires a little performance. You take it out, orient it, hold it towards something and maintain that position. There is enough choreography that people nearby can infer what is happening, even if they can’t know for certain whether they are the subject.
Glasses collapse much of that choreography. Recording becomes easier, more ambient and less legible to the people around the wearer. The important change may therefore be less about what the technology can do and more about how difficult it is for another person to know that it is happening.
That uncertainty deserves serious consideration. But if we are going to have a useful privacy debate, “someone could record me without asking” is not nearly precise enough. Someone could already do that.
The more useful question is what kinds of observation and recording we believe should require consent, in what circumstances, and why.
Privacy where?
The current discussion treats “privacy” as though it describes one obvious harm. But privacy in a bedroom, privacy gardening in your backyard, and privacy while walking through a shopping centre are not the same thing. The word is understood; what isn’t being articulated is exactly what people fear smart glasses will change.
Some situations carry a strong and widely understood expectation of privacy. Using a bathroom, getting changed, having a medical consultation, or talking with someone inside a private home all sit comfortably in that category. But the boundaries become less obvious as we move into semi-private and public spaces, where being visible to other people does not necessarily mean being comfortable with being recorded, identified, analysed or published.
Australian law already makes some of these contextual distinctions. State and territory surveillance laws regulate private activities and conversations in different ways. Since 2025, Australian law has also allowed people to bring a civil claim in court for a serious invasion of privacy in certain circumstances.
Most of us already accept that we will sometimes appear incidentally in somebody else’s photograph or video in public. If you spend enough time standing in front of Flinders Street Station in Melbourne, or in front of the Sydney Opera House, there is an excellent chance that the back of your head has already appeared in someone’s holiday album.
The situation feels different when you become the deliberate subject of a recording. Different again when you do not know it is happening. And very different when that recording is edited, mocked, sexualised, monetised, or broadcast to thousands of strangers without context.
What we call a privacy problem may actually be a collection of quite different harms.
Covert recording can itself be profoundly harmful. In stalking, domestic and family violence, harassment or other targeted abuse, secretly observing someone may be part of a much larger pattern of monitoring and control. The fact that cameras have existed in other forms for decades does not make a device that makes covert observation easier, benign. Tech-facilitated abuse already shows how everyday consumer devices can acquire a very different significance when they are used by someone intent on controlling another person.
But recording can also be the beginning of something else.
Increasingly, smart glasses can interpret what they see: read text, recognise objects, describe surroundings, identify people, and potentially connect visual information with other data.
Incidentally, this is also one of the most significant things about the technology for me personally. The ability to have the world described, objects identified, text read and visual information interpreted in real time, can give me access to things I would otherwise miss entirely.
Some of those capabilities are extraordinarily useful. Some are far more intrusive.
Can we reconcile those interests well enough to protect people from harm without stripping away the things this technology can genuinely make possible?
Looking has always been unequal
For me, the camera is not primarily interesting because I can record what is around me. It is interesting because the technology can interpret what is around me.
I already have image recognition on my phone. I can point it towards an object, piece of text or scene and ask for information. The capability is not entirely new.
What changes when that technology moves onto my face is the relationship between me and the environment.
Using a phone turns access to visual information into a task. I have to find the phone, unlock it, open the right application, point the camera towards the thing I cannot see properly, attempt to maintain the framing, and then hope I have produced a usable image.
With glasses, I can look towards something and ask.
That difference sounds small when described as “hands-free”. In practice, it can be the difference between using a tool occasionally and having information available as part of moving through the world.
It also raises a harder question when we debate whether smart glasses should have cameras:
What are people ordinarily entitled to know when they are physically present?
Sighted people continuously collect and interpret information about the people and places around them. They see who entered a room, what someone is wearing, whether the seat beside them is free, what is written on a person’s name badge and whether someone across the room is holding a phone camera in their direction.
Nobody asks them to obtain consent before gathering most of this information. We simply call it seeing.
Give technology the job of translating some of that same visual information for a blind person and the boundaries suddenly become much harder to define.
I am not arguing that whatever a sighted person can see, a machine should therefore be allowed to identify, analyse, store and cross-reference indefinitely. That would be an impressive leap from equality into dystopia.
But I do think we need to work out where looking ends and surveillance begins.
A pair of glasses may be able to tell me that someone has entered the room. Should they tell me who the person is? They may be able to read a slogan on someone’s shirt. Should they search for that person online? They might identify that someone is holding a phone towards me. Would warning me that I may be being recorded actually increase my privacy?
Those questions don’t have one obvious answer, because the interests involved are not always in opposition and they are not always equal.
A useful privacy framework cannot begin and end with whether the camera happens to be worn on someone’s face.
The strange privilege of knowing you are being watched
One part of the smart-glasses debate keeps returning to visible recording indicators: a light turns on, an icon appears, the hardware signals that recording has begun.
That sounds reassuring until you ask who can actually perceive the signal.
A tiny light may satisfy a product requirement without providing meaningful notice to everyone around the wearer. The same problem already exists with surveillance cameras accompanied by small signs at the entrance, surrounded by lots of other information. Disclosure has technically occurred, but whether anyone has received or understood the information is a different question.
We often treat visibility as though it gives people agency. If you can see the phone being raised, you can move, object, ask what is happening, or change what you say. If you cannot perceive that cue, none of those choices are available to you.
The design challenge is to make notification meaningful enough that everyone can actually respond to it.
Otherwise, we risk creating privacy protections that work best for the people already most able to detect that they need them.
There is also a difference between something being legal, something being socially acceptable, and something simply feeling creepy. Smart glasses make those boundaries harder to read because the behaviour itself is less visible. Is the person recording? Using visual assistance? Taking a photograph? Asking the device to read a sign? Doing nothing at all?
The task is to work out what meaningful notice, reasonable expectations, and appropriate safeguards look like when the technology no longer announces itself in familiar ways.
Recording can redistribute power too
The privacy discussion also tends to assume that the person with the camera always has more power.
Sometimes they do. Sometimes recording gives power to a person who previously had very little.
I know this from something as ordinary as trying to get into a taxi with my guide dog. Refusals still happen regularly and, when they do, the complaints process almost assumes I will have some kind of recorded evidence.
So, I often approach the interaction with my camera already running.
I hate that.
I don’t want every taxi ride to begin as evidence collection.
Nor do I want to assume the driver will do the wrong thing, and I am genuinely delighted when the recording becomes completely useless. But I also know how quickly an experience can become difficult to establish once it enters a formal complaints process without documentation.
Video can provide evidence of discrimination, harassment, abuse, threats or behaviour that would otherwise become one person’s account against another. For people whose accounts are already less likely to be believed, retaining evidence can materially change the balance of power.
There is a reason dashcams have become commonplace, people record encounters with authorities, employees save emails, and consumers take screenshots before lodging complaints. We have built systems in which documentation is often treated as more authoritative than the person describing what happened.
That creates an uncomfortable tension. More ubiquitous recording may make some harmful behaviour easier to evidence and, perhaps, make people more accountable for how they behave when they believe someone is watching. There is something worth exploring in the old idea that character is revealed by what we do when nobody is looking.
But it would be far too convenient to conclude that people worried about smart glasses are simply afraid of accountability.
A woman should not have to accept covert footage of her body because she went to the beach. A child cannot meaningfully consent to becoming online content. Someone disclosing personal information in what they believed was a private conversation has a legitimate interest in where that information goes.
“Nothing to hide, nothing to fear” has always been a particularly poor way of thinking about privacy. Privacy gives us room to be unfinished, contradictory, vulnerable, foolish and human without every moment becoming evidence.
I am, for example, deeply grateful that the ubiquity of social media arrived after my twenties. I see no compelling public interest in correcting that historical evidence gap.
The law already asks a more useful question
The legal position becomes muddled because Australians often talk about “the Privacy Act” as though it gives everyone a general right not to be photographed without consent.
It doesn’t work that way.
The Privacy Act 1988 primarily regulates how Australian Government agencies, businesses and certain other organisations handle personal information. A private person taking a photograph or video for personal purposes will likely sit outside those information-handling rules.
That doesn’t mean anything goes, either. State and territory surveillance laws regulate private conversations and activities in different ways, while criminal laws may also apply to stalking, harassment, voyeurism and intimate images.
The relatively new legal avenue for serious invasions of privacy is significant because Australian privacy law has historically been much better at regulating how organisations handle information than giving individuals a general remedy when another person seriously intrudes on their privacy.
It still isn’t a general right not to be photographed. Nor does being in public mean every possible use of information about you is automatically acceptable.
The legal picture is messier than either of those propositions, but perhaps usefully so. It recognises that the circumstances surrounding an intrusion are part of deciding what kind of privacy harm has actually occurred.
The tech company is in the interaction too
So far, the debate mostly positions two people against each other: the person wearing the glasses and the person potentially being recorded.
But there is a third participant in the interaction: the company behind the technology.
The wearer is not necessarily the only one receiving whatever the glasses see. Images may be processed locally on the device, sent elsewhere for analysis, retained for some period, used to improve services, or connected with other information.
If I ask my glasses whether the approaching tram is an 86, there is a considerable difference between a system briefly analysing the scene and discarding it, and one retaining identifiable footage of everyone waiting at the stop.
The people at the tram stop never bought the product. They did not agree to its terms and conditions, and they may have no idea that information about them has entered someone else’s technology ecosystem.
That raises a different privacy question from whether I personally pressed record.
We spend a lot of time asking what obligations someone wearing smart glasses has to the people around them. We should probably spend at least as much time asking what obligations the companies behind those glasses have to people who never chose to interact with their product at all.
Perhaps cheap glasses are doing us a favour
Part of the urgency appears to have come from camera glasses becoming cheap enough to be ordinary. There is something psychologically different about technology once it turns up at Kmart.
A capability that feels futuristic when it costs hundreds or thousands of dollars suddenly becomes an immediate social concern when somebody can buy it alongside a throw rug and pairs of thongs.
That reaction is understandable. Cheap technology scales both benefits, and misuse.
But perhaps inexpensive smart glasses are also forcing us to have a privacy conversation we were going to need anyway.
Cameras will continue to become smaller. AI will become better at interpreting what cameras see. Wearable computing will not end with this generation of glasses. The next device may sit in a pendant, watch or clothing.
If our privacy framework depends on recognising a particular object as a camera, we are going to spend the next decade chasing hardware.
The more durable questions concern what the technology is doing with the information and what harm that has the potential to create.
Those questions still work when the camera changes shape.
I want the glasses
If you haven’t figured it out already, I want this technology to improve.
I want to be able to walk into a room and know more about what is there. I want to access visual information without awkwardly placing a phone between myself and the world. I want technology to give me some of the information that other people acquire effortlessly by looking around.
I would also quite like not to be secretly recorded in a bathroom.
These positions are not incompatible.
We have a habit in accessibility debates of constructing rights as competitors: one person’s access against another person’s privacy, one person’s independence against another person’s safety or comfort. It makes for excellent conflict, and poor design.
The more ambitious challenge is to create technology, norms, and regulation capable of holding more than one legitimate human interest at once.
The technology will need social norms around it, just as mobile phones did. People will learn what feels acceptable, what feels intrusive, and when somebody probably needs to explain what their glasses are doing.
Those norms will develop whether we deliberately shape them or not. It would be better if they were informed by both the people who may be harmed by the technology and the people whose independence it can increase.
Perhaps smart glasses are revealing the privacy problem
I started thinking about this because so much of the discussion around smart glasses has felt frustratingly simplistic. The technology is described as either exciting innovation or an obvious privacy threat, with surprisingly little room for both things to be true.
They are true.
Smart glasses can make covert recording easier and create serious risks in the hands of someone intent on stalking, controlling, exploiting or humiliating another person. They can also give me access to visual information that I would otherwise miss and make parts of everyday life more independent.
I have an obvious self-interest in getting this right. I want the technology.
But I also want us to be capable of having a more sensible public debate than deciding that one legitimate concern must defeat another.
That becomes more important as restrictions begin to accumulate. Smart glasses are already being banned or limited in a growing range of places and events around the world. Some of those restrictions may be entirely appropriate. My concern is not that nobody should ever ban them; it is what happens when prohibition becomes the default response before we have properly distinguished the behaviour we want to prevent from the uses we want to preserve.
If privacy advocates are not listening to people whose independence can be increased by this technology, the response will be incomplete. If accessibility advocates and technology companies are not listening to people at risk of stalking, harassment, coercive control or exploitation, their response will be incomplete too.
We should be able to hold both sets of knowledge at once.
That is partly why this debate interests me beyond the glasses themselves. New technology will continue to present us with uncomfortable combinations of benefit and harm. If our response is to choose the most emotionally compelling side and stop listening, we will become very good at making fast decisions and much less good at making wise ones.
I want technology that helps me understand more of the world around me. I also want other people to have meaningful privacy, safety and control over their own lives.
My self-interest makes me want us to get this particular debate right.
My broader interest is in living in a society capable of listening to all of the people affected before deciding what progress should look like.
