Trends — Field Notes on Ambient Technology
The Consent Recession
Smart hardware finally works. That was never the hard part. The hard part is that the most important person in the room is no longer the one wearing the device.
12 August 2026 · 9 min read
For a decade, the smart hardware industry asked a single question: can we build it? Can a camera disappear into a spectacle frame. Can a thermostat infer that you are cold before you say so. Can a processor small enough to sit in an ear canal translate Portuguese into English fast enough to hold a conversation.
The answer, in 2026, is yes to all three. Smart glasses shipments grew 139% year over year in the second half of 2025, and AI glasses made up 88% of them. Meta says it has sold seven million pairs. The smart home market is projected at $207 billion this year. Hearables are forecast to ship 407.6 million units in 2026.
The engineering question is settled. What replaced it is harder, and the industry is nowhere near answering it: when a device becomes invisible, what happens to the consent of everyone who did not buy it?
“There are some pretty dark places we could go here. I’m not anti-technology in any sense, but as a societal matter — will I need to think about being recorded anytime I go out in public?”
David Kessler, US Privacy Practice Head, Norton Rose Fulbright — to the BBC, May 2026
I. The Asymmetry
The Buyer Consents. The Bystander Cannot.
Every previous consumer device negotiated its social contract in the open. A camera was visibly a camera. A phone raised to eye level announced itself. The person being photographed could see the act and object to it, or move out of frame. That visibility was not a design flaw — it was the mechanism by which public space stayed liveable.
AI glasses remove that mechanism. Meta’s design places an almost invisible camera in a Ray-Ban frame; recording starts with a touch. There is an LED indicator, but as one daily wearer told the BBC, it “appears dim in daylight and often goes unnoticed” — most people assume he is wearing ordinary eyeglasses.
The consequence is not hypothetical. Women in the UK have been approached and filmed without their knowledge, discovering the footage only when it circulated online with abuse attached. One woman asked the poster to take a recording down and was told removal was “a paid service.” Photography in public is broadly legal; her legal recourse was close to nothing.
This is the asymmetry that defines the category. The buyer performs an act of consent at the point of purchase. Everyone else within the field of view is enrolled without being asked, without being told, and — increasingly — without being able to tell.
The indicator light is the entire social contract. In daylight, it is close to invisible.
II. The Correction
Institutions Move Before Regulators Do
When the law is slow and the technology is fast, the correction arrives through the door policy. That is exactly what is happening now, and it happened within weeks.
In early August 2026, the Guardian reported that venues across the UK — from elite restaurants to popular pub chains — had banned the £359 glasses to protect the privacy of customers and staff. Days later, courts in England and Wales barred them from judicial buildings entirely: anyone attempting to enter wearing a pair has them confiscated and returned on exit, with the court service citing unauthorised recording and contempt risk.
Read that sequence carefully, because it is the most important signal in the category this year. Restaurants, theatres and courthouses are not privacy activists. They are risk managers. When they move faster than legislators, it means the liability has already become concrete — and it means the product is being priced out of specific physical spaces one venue at a time.
Berkeley’s David Harris, a former Meta AI researcher, expects this generation to hit the wall that killed Google Glass in 2013: “Technology like this is fundamentally an invasion of privacy and it’s really going to face more and more backlash.” Meta’s counter-position, from CTO Andrew Bosworth, is that sales volume itself proves acceptance. Both cannot be right.
III. The Same Pattern, Quieter
Your House Has the Same Problem. It Just Doesn’t Leave the Building.
The glasses debate is loud because the harm is visible and the victim is a stranger. The smart home version is quiet because the surveillance is domestic, consensual at purchase, and comfortable. It is also far larger: a $207 billion market in 2026, up from $162.8 billion in 2025.
But the structure is identical. The person who bought the doorbell camera consented. The delivery driver did not. The houseguest sleeping in the spare room did not consent to a presence sensor logging when they came in. The cleaner did not agree to be timestamped. Every smart home is a small jurisdiction whose laws are written by one household member and applied to everyone who walks in.
The difference is that a home is bounded. You can ask. You can decline to enter. Public space offers no equivalent exit — which is precisely why the glasses hit the wall first, and why the home category has been allowed to grow ten times larger without comparable scrutiny.
IV. The Exception
Why Earbuds Got a Pass
Hearables are the largest of the three categories by volume — 407.6 million units forecast for 2026, growing 4.0% year over year — and they carry microphones, health sensors and live translation. By the logic above, they should be the most contested product on the market. They are the least.
The reason is instructive: earbuds are visibly worn but incapable of capturing you. They process the world inward — noise cancellation, hearing assistance, translating what reaches the wearer’s ear. Nothing points outward at a stranger’s face. When Apple shipped live translation and heart-rate sensing in AirPods Pro 3, the discourse was about accuracy and battery life, not about the people standing nearby.
This is the cleanest natural experiment the industry has. Three categories, comparable AI capability, comparable intimacy, wildly different levels of backlash. The variable that predicts trouble is not intelligence, sensing, or data collection. It is whether the device can capture a person who never agreed to be captured.
The Bystander Test
Three questions that predict whether a smart device will face structural backlash. They have nothing to do with specifications.
Can it capture a non-owner?
If yes, the product has a constituency it never sold to.
Can that person tell?
An indicator nobody notices is not disclosure. It is documentation.
Can they opt out without leaving?
In a home, yes. On a street, no. That gap is where regulation lands.
V. What Comes Next
The Next Eighteen Months
Apple is reportedly developing its own glasses. Snap has committed to new Specs. Google is returning to the category more than a decade after pulling Glass. Researchers cited by the BBC expect as many as 100 million people to buy a pair within a few years.
If that happens, the enforcement problem stops being about one company’s product decisions. Courthouses, hospitals, theatres, schools and bathrooms all operate on rules that assume cameras are identifiable objects. At 100 million units, that assumption fails everywhere at once — and institutions will not wait for a regulatory framework before writing door policies.
The reported plan to add facial recognition to a future version is the fault line to watch. Covert recording is a privacy problem. Covert recording plus instant identification is a categorically different one, and it is the specific capability the Electronic Frontier Foundation has argued would “obliterate the privacy of everyone.”
Our position is narrow and, we think, defensible: this category will not be decided by camera resolution, battery life, or AI model quality. It will be decided by whether manufacturers build disclosure that actually works on a sunny afternoon — and whether they do it before the door policies harden into law.
The machines finally work. The question is no longer what they can do for the person who bought them — it is what they do to the person standing in frame.
Research Appendix
Every figure in this piece is listed below with its source and publication date. Where estimates from different research firms diverge, we cite the specific firm rather than presenting a consensus that does not exist.
| Smart glasses shipments +139% YoY, H2 2025; AI glasses = 88% of shipments; Meta >80% share | Counterpoint Research, 26 Feb 2026 — Global Smart Glasses Shipments Grew 139% YoY in H2 2025 |
| 7 million pairs sold; indicator LED dim in daylight; Kenya annotation lawsuits; Kessler and Harris quotes; 100 million projected buyers; facial recognition plans | BBC News, 13 May 2026 — Smart glasses are ‘an invasion of privacy’ — Meta’s are selling better than ever |
| UK restaurants, pubs and theatres ban £359 glasses | The Guardian, 6 Aug 2026 — Restaurants, pubs and theatres ban Meta’s ‘spy glasses’ over privacy fears |
| Courts in England and Wales confiscate glasses at entry | The Guardian / AOL UK, 12 Aug 2026 — Meta glasses banned from courts in England and Wales |
| Covert filming of women; removal quoted as “a paid service” | BBC News, 2026 — Woman covertly filmed for ‘humiliating’ social media content — then told to pay |
| Facial recognition on smart glasses would “obliterate the privacy of everyone” | Electronic Frontier Foundation, 10 Mar 2026 — Think Twice Before Buying or Using Meta’s Ray-Bans |
| Smart home market $162.8B (2025) → $207.0B (2026) | Grand View Research — Smart Home Market Size, Share & Growth Report 2026–2033 |
| Hearables 407.6M units in 2026, +4.0% YoY | IDC Worldwide Wearables Tracker, 2 Jul 2026 — Wearable Devices Market Insights |
| Live translation and heart-rate sensing in AirPods Pro 3 | TechCrunch, 9 Sep 2025 — AirPods Pro 3 arrive with heart-rate sensing and live translation |
| 2025 smart glasses shipments: 7.25 million units | Treeview XR Market Statistics, 1 Aug 2026 — XR & Smart Glasses Market Statistics Report |
Note on method: market sizing for smart home and hearables varies materially between research firms — estimates for the 2026 smart home market range from roughly $159B to $231B depending on category definitions. We cite Grand View Research and IDC consistently rather than blending incompatible methodologies. Meta’s seven-million figure is company-reported and has not been independently audited.
Leave a comment