Meta Smart Glasses, Invisible Cameras and the Right to Your Own Face

Diane Morgan was working. The British comedian, best known as the gloriously ill-informed Philomena Cunk, was filming on Hollywood's Walk of Fame when a man approached her wearing Meta Ray-Ban smart glasses. He asked for her telephone number, told her she was gorgeous and, when a crew member intervened, reportedly became threatening.

It might have remained an unpleasant encounter between a woman trying to work and a man refusing to leave her alone. Except he was filming it.

The resulting content was posted online and attracted more than a million views.

Sandeep De recently drew attention to the incident on LinkedIn, following reporting by Futurism. But he also pointed to something considerably more troubling: this was not simply an obnoxious person who happened to own a camera. The camera was part of the encounter.

The Camera You Don't See

We have lived with cameras in public for decades. Most of us understand the social grammar of being photographed. Someone takes out a phone. They raise it. They point it at us. Perhaps they ask permission; perhaps they don't. But we normally know that the relationship between observer and observed has changed.

That tiny piece of information matters.

It allows me to turn away. To leave. To object. To stop talking. To cover my face. To ask why you are filming me. Or simply to behave differently because I now know I have an audience.

Meta's Ray-Ban smart glasses fundamentally alter that social grammar. The recording device is incorporated into an ordinary-looking object worn on the face. Meta points to a recording indicator light, but the existence of an LED does not answer the more fundamental question of whether the people being captured meaningfully understand that recording is occurring. And reports suggest users have found the obvious low-tech response to the indicator: covering it.

The important technological innovation isn't merely wearable recording. It is recording without the traditional social signal of recording.

That changes the relationship between the person operating the camera and everybody around them.

Now Put Them in a School

Timothy Cook recently highlighted an even more disturbing manifestation: teenage boys reportedly using Meta glasses in schools to record girls without their agreement and upload the resulting videos to TikTok.

Here, the technology collides with another boundary. A school corridor is not simply another public street. Schools already regulate phones, photography, safeguarding, harassment and student behavior because children exist within an environment in which the institution assumes responsibilities that do not apply to two strangers passing one another on a pavement.

Yet smart glasses can circumvent precisely those controls.

Cook describes one student apparently putting his phone into a required confiscation pouch while continuing to wear the device capable of recording. There is something almost perfect about that image.

We built the rule around the phone. Technology moved the camera onto the child's face. The policy survived. Its purpose didn't. And this is where treating smart glasses simply as another privacy problem becomes inadequate.

What Exactly Has Been Taken?

Suppose someone secretly records you. What has actually been captured?

• Your face.
• Your voice.
• Your gestures.
• Your expressions.
• Your movements.
• Your reactions.
• Your interactions with other people.
• Your location and context.
• Perhaps your distinctive mannerisms and patterns of speech.

Twenty years ago, the obvious answer would have been: a video of you. That answer is no longer sufficient. Because a contemporary recording can become data.

It can be copied, indexed, searched, analyzed, edited and distributed. Facial-recognition technology can identify the subject. Voice technologies can isolate speech. Generative AI can manipulate the recording or potentially use captured attributes in systems capable of producing synthetic representations.

The recording is not necessarily the final product. It may be raw material for another version of you.

This is precisely why the question “how do I protect my digital identity?” is becoming very different from the older question “how do I stop someone publishing my photograph?” People increasingly want to know how to stop AI from using their likeness, protect against deepfakes, prevent voice cloning, remove impersonation videos and control their digital likeness. Those sound like different problems because our legal system puts them into different boxes. From the perspective of the human being, they may be one problem.

This problem is not entirely new. Ten years ago, my former Icondia colleague Keith Laker wrote Your Face is Big Data – and valuable!, prompted by Egor Tsvetkov's experiments showing how facial recognition could connect photographs of apparently anonymous strangers with their online identities. Keith saw immediately that this was not merely a privacy problem. If the face was becoming commercially valuable data, he argued, perhaps the individual should possess a legal mechanism for controlling—and potentially benefiting from—its commercial exploitation.

A decade later, the question has become considerably more complicated. Facial recognition asked: who is this person? Generative AI can increasingly ask: how can I reproduce this person?

Privacy Law Sees Data. Personality Law Sees the Person.

Consider what happens legally after the recording.

• We might ask whether the subject possessed a reasonable expectation of privacy.
• Data-protection lawyers may ask whether personal data was processed and what lawful basis applied.
• If the conduct becomes persistent, threatening or targeted, harassment law may become relevant.
• If the resulting material damages reputation, defamation may enter the picture.
• If it creates the appearance of endorsement, trademark law or passing off may sometimes help.
• If existing protected material has been copied, copyright may provide another remedy.

Each question is legitimate. But notice what has happened.

The law has disassembled the person.

• Face becomes biometric data.
• Voice becomes a recording.
• Reputation becomes a defamation interest.
• Commercial identity becomes goodwill.
• Creative material becomes copyright.
• Name or recognizable commercial identifiers may become trademarks.
• Privacy becomes information or circumstances in which the individual possesses a reasonable expectation of privacy.

Yet the person experiencing the intrusion does not experience herself in pieces. She experiences somebody taking her.

That is the gap I have been exploring in my work on personality rights and, increasingly, artificial intelligence.

The Right to Your Own Face

This becomes particularly uncomfortable in Britain because English law remains reluctant to recognize a general image or personality right. There are good historical and jurisprudential reasons for that reluctance. English law quite properly worries about freedom of expression, journalism, parody and the consequences of giving people broad proprietary control over how they may be represented.

But the absence of a general right produces an increasingly peculiar result.

• English law may protect your privacy.
• It may protect your personal data.
• It may protect you against false endorsement.
• It may protect copyrighted works involving you.

What it generally does not say is something intuitively much simpler: This is your recognizable personality, and another person does not acquire unrestricted authority over it merely because technology enabled them to capture it.

AI makes that distinction increasingly important.

As I have argued previously in writing about image rights in the age of AI, modern identity can be replicated without exact copying, synthesized without direct appropriation and commercialized without clear attribution. Existing legal tools consequently remain reactive, fragmented and jurisdictionally inconsistent.

Smart glasses move that problem one stage earlier.

Before we can ask whether AI is allowed to reproduce you, we have to ask how machines acquire the manifestations from which digital identity can be constructed in the first place.

The Woman on the Pavement Is Not Content

There is another dimension to the Diane Morgan incident that shouldn't disappear beneath legal doctrine. The economics matter.

• A stranger's discomfort can become somebody else's content.
• Her anger can become engagement.
• Her resistance can become entertainment.
• An altercation can generate views.

Platforms have constructed extraordinarily effective markets for human attention, and certain creators have learned that provocation produces attention very efficiently. Wearable cameras lower the friction involved in producing that content.

Women and girls are particularly vulnerable where the genre depends upon unwanted approaches, sexual attention or provoking reactions. In those circumstances, telling the subject that she is “in public” rather misses the point.

She isn't objecting to being seen. She is objecting to being appropriated as content.

Those are not the same thing.

Children Should Not Have to Become Surveillance Experts

The school examples make the absurdity clearest. We cannot reasonably tell a teenage girl that the solution to invisible recording is permanent vigilance.

• Check everyone's glasses.
• Look for an LED.
• Work out whether it has been covered.
• Assume every interaction might be streamed or recorded.
• Learn the platform reporting mechanisms.
• Understand whether your state requires one-party or all-party consent.
• Work out whether the resulting video constitutes harassment, an invasion of privacy or unlawful processing.
• And, increasingly, perhaps learn how to find AI-generated clones of yourself and get unauthorized content removed.

That is not meaningful protection.

It is the transfer of technological risk from the people who create and deploy technology to the people most vulnerable to its misuse. Children should be learning how to participate in society. They should not need to become counter-surveillance specialists simply to walk between classrooms.

Consent Cannot Be Reduced to a Blinking Light

This is ultimately why I am uncomfortable with technical discussions that focus excessively on whether smart glasses display a recording indicator. The light matters. But it does not solve the problem.

Consent is not a hardware feature.

Meaningful human consent requires some combination of knowledge, comprehension and genuine opportunity to choose. A technological system designed around persistent wearable capture, therefore, creates a much broader governance question.

• What responsibility belongs to the user?
• What responsibility belongs to the school or workplace?
• What responsibility belongs to TikTok, Instagram or another platform distributing the resulting material?
• And what responsibility belongs to the manufacturer that designed the recording architecture?

Those questions cannot all be answered by pointing to an LED.

From Privacy to Personality

This is why I increasingly believe we need to widen the conversation from privacy to personality.

• Privacy asks an essential question: Was I entitled to be left alone?
• Personality asks another: What authority should I retain over recognizable manifestations of myself once they have been externalized?

That includes face, voice, gesture, mannerism, appearance and other attributes through which a human being becomes recognizable. Artificial intelligence makes that second question unavoidable because technology can now detach those manifestations from the human being who produced them and create entirely new representations.

My broader work on personality law has moved toward treating personality not merely as privacy, property, reputation, data or intellectual property, but as the continuing legal relationship between a person and the external manifestations through which that person becomes socially, culturally and economically recognizable.

Smart glasses are an unexpectedly good illustration of why that matters. They sit at the boundary between physical and digital identity.

• They capture the physical person.
• They convert manifestations of that person into digital information.
• Platforms distribute them.
• AI can increasingly manipulate them.

And throughout that chain, the law asks different questions because each stage falls into a different doctrinal category. The human being remains the same.

No. That Part of Me Is Not Yours.

Technology has always changed social expectations. Cameras became ubiquitous. Phones became cameras. Cameras became connected to global publishing platforms. Now the camera is disappearing into things we wear. The next step will not stop there.

Wearable AI will increasingly see what we see, hear what we hear and interpret the people around us. The debate about Meta glasses is not really about one pair of spectacles. It is about the rules we establish for a world of persistent machine observation.

There will be legitimate uses for that technology. There will be enormously valuable ones. Accessibility alone provides compelling examples of what wearable vision systems may eventually accomplish. But technological usefulness does not extinguish the rights of everybody standing within sensor range.

Nor should being visible mean being available for capture, distribution, analysis, monetization or synthetic reproduction. That distinction becomes particularly important as the boundary between recording a person and acquiring data about a personality begins to disappear.

The question is no longer simply: Were you entitled to film me?

Increasingly it is: What are you entitled to do with the version of me you captured?

And ultimately there needs to be a point at which the human being can answer: No. That part of me is not yours to take.