Oslo has barred camera glasses from every school day after its education chief said children were being turned into unpaid extras in a Silicon Valley recording business. The rule is already in force. The national government is still writing the study group. That gap is the story.
 |
| Source: Video Screenshot |
Key Takeaways by Planet Today
Local action versus national delay: Oslo’s ban took effect on 27 August 2026 and was announced on 10 September after a political pause for the death of King Harald V. Bærum followed for grades 1–10. The Labour-led government has asked municipalities to write the devices into school rules and has set up an expert group, but has not passed a nationwide product ban.
The product, not the slogan: Meta’s Ray-Ban and Oakley smart glasses can photograph, film, take calls, translate text, overlay maps and talk to an on-device AI assistant. The city’s charge is not that every wearer is a spy. It is that the hardware is built so bystanders cannot reliably know when a camera is live.
What Meta actually changed this month: After reports of LED tampering, Meta disabled cameras on thousands of altered pairs and shipped software that stops recording if the capture light is covered mid-take. The company calls the glasses “privacy-forward” compared with a phone. Critics say a blinking light on a face is not consent.
The facial-recognition shadow: Reporting in early 2026 described an internal Meta feature nicknamed “Name Tag.” Civil-liberty groups called it a red line. Code linked to face recognition later appeared and disappeared from a companion app. Oslo’s text treats that capability as close enough to regulate now.
Why a schoolyard rule does not settle the market: A municipal code of conduct can keep glasses out of classrooms. It cannot stop sales in Norwegian shops, use on the tram, or recordings of pupils on the walk home. That is why Oslo is pressing the cabinet, and why other European capitals are watching the same file.
What Oslo Said, and What It Did Not Wait For
On Thursday, 10 September 2026, the City of Oslo confirmed a ban that had already been running for two weeks. Smart glasses are barred for the full school day, at every level of municipal education. The phrasing that travelled with the AFP wire was blunt. Vice Mayor for Education Julie Remen Midtgarden, a Conservative, said politicians needed “the courage to be honest about what this is really about: one of the world’s richest companies has launched a product whose entire business idea is to make it impossible to know when you are being filmed.” Then the line that became the headline: “In Oslo we refuse to let our children become extras in Meta’s data collection.”
Norwegian readers saw a slightly fuller version in Aftenposten. Midtgarden added that the glasses “are being sold in Norway today, they are being used in Norwegian school playgrounds today, and the problem is here now.” An expert group, she said, “is not an answer to parents who wonder whether their children are being filmed in secret.” She also said a local code of conduct is not enough. Regulation beyond the school gate is a national job. “Oslo has taken responsibility. Now it is the government’s turn.”
The announcement was late on purpose. The ban itself started on 27 August. King Harald V died the next morning. Norwegian parties observed a political truce through the mourning period and the state funeral on 9 September. The city waited one more day, then published. That sequence is not a conspiracy. It is how Oslo chose to handle a commercial product during a royal death. It does, however, explain why international wires landed on 10 September as if the rule were new.
Bærum, the large suburban municipality west of the capital, moved in the same week. Its notice treats AI glasses as private digital equipment, banned at school and in after-school care on the same footing as phones and smartwatches, for pupils in years 1 to 10. Two neighbouring councils, two slightly different legal hooks, one practical result: the devices stay in the bag or at home.
How the National Government Has Played the Same File
Oslo’s Conservative city hall is not inventing a quarrel with a silent cabinet. It is accelerating a quarrel that already existed inside the Labour-led government.
In mid-August, Education Minister Kari Nessa Nordtun told TV 2 she did not want a statutory ban on teachers using the glasses in primary school. She said most teachers could be trusted and that she would look again if “unwanted incidents” appeared. By late August the same ministry was asking municipalities and schools to write Meta glasses and other AI eyewear into their conduct rules. Those two positions can sit in one file if you treat classroom culture as local and product law as national. Midtgarden’s point is that parents do not experience the file that way. They experience a playground.
Digitalisation Minister Karianne Tung has taken a harder line on one specific feature. In late August she told VG she wanted stricter rules for smart glasses and was prepared to consider a ban on facial recognition of other people in public space. Forbrukerrådet, the consumer council, and Datatilsynet, the data-protection authority, have already argued that sales themselves should be halted pending a proper review. The government answer so far is an expert group and a promise to study how society can be protected. That is the sentence Oslo keeps quoting back.
None of this is unique to Norway. It is the Nordic version of a wider European argument: who moves first when a consumer gadget arrives faster than a statute. The same week Oslo spoke, AFP reported that a British petition was calling for a sales ban, that an EU lawmaker had asked the Commission how the glasses sit with data-protection and AI rules, and that the European Data Protection Board had commissioned a report on “social acceptability,” due in the autumn. France’s CNIL has told the public to stay vigilant. The Commission’s reply, for now, is that national data-protection authorities already hold the enforcement job. That answer is legally tidy. It is also why a city council in Oslo can look faster than Brussels.
What the Glasses Actually Do
Strip the slogans and the hardware is not mysterious. Current Meta smart glasses, made with EssilorLuxottica under the Ray-Ban and Oakley brands, look like ordinary frames. They carry cameras, microphones, speakers, wireless links and an AI assistant. A wearer can take a still or a clip, play music, take a call, have a foreign sign read aloud, see text or a map in the field of view, and ask the assistant what is in front of them. Meta sold about seven million pairs in 2025, after roughly two million across 2023 and 2024 combined. Google’s first Glass wave failed in 2013. This wave did not.
The privacy fight is not about translation or music. It is about a camera that sits at eye height, points where a person looks, and does not look like a camera. Meta’s designed answer is a capture LED that is supposed to stay on whenever the device is recording. In 2026 the company has spent the summer closing holes in that answer. After users learned to start a recording and then cover the light, Meta shipped an update that kills the camera if the LED is blocked mid-take. After finding pairs with the light drilled out or taped over, it disabled cameras on thousands of devices — fewer than one tenth of one percent of units sold, the company says — and left those buyers with expensive headphones. Alex Himel, Meta’s vice president for wearables, has called the contest a “cat-and-mouse game” that will not end. Andrew Bosworth, the chief technology officer, has argued the other way: a phone has no recording light at all, and the phone is the most common camera on earth.
Both statements can be true at once. A light is more notice than a black slab in a pocket. A light on a pair of sunglasses is still easy to miss in a crowded schoolyard, especially if the child being filmed is eight years old and looking at a football. Oslo’s sentence — “impossible to know when you are being filmed” — is a political compression of that gap. It is not a laboratory finding. It is a judgement that the remaining uncertainty is too high for a place full of minors who cannot contract, cannot opt out of a classmate’s gadget, and cannot later unsay a clip that left the building.
The same logic has already reached other institutions. Private schools in England have written the glasses out of their grounds. Districts in Florida, California, Nevada and Cambridge, Massachusetts, have done the same. New York state courts barred recording eyewear from more than a thousand courthouses in July. UK cinemas moved against them over piracy. U.S. Immigration and Customs Enforcement told staff not to wear the consumer pairs at work, while the Department of Homeland Security had, earlier in the year, asked for money to prototype official glasses that could support biometric identification. Those two federal facts sit beside each other without cancelling each other. They do show why a Norwegian city talking about “Meta’s data collection” is not speaking a private dialect.
The Facial-Recognition File That Will Not Stay Closed
Oslo’s statement said Meta had “reportedly developed a facial recognition feature ready to be integrated into its smart glasses.” That sentence needs the reporting trail attached, because it is the load-bearing claim for people who think the city is overreaching.
In February 2026 the New York Times described an internal project known as “Name Tag.” Wearers would identify people in view and pull information through Meta’s assistant. Engineers discussed a narrow version limited to existing Meta connections and a wider version that could reach anyone with a public Meta account. Mark Zuckerberg, according to people briefed on the plans, wanted the feature because it would make the assistant more useful and the product harder to copy. The company had looked at face recognition for the first Ray-Ban Metas in 2021 and pulled back. It looked again as sales rose.
In April more than seventy organisations, including the ACLU, EPIC and Fight for the Future, told Zuckerberg to kill the idea. Their letter said design tweaks and opt-outs would not fix a camera that can name a stranger in the street. They warned about stalking, domestic abuse, immigration enforcement and the targeting of people who had never bought the product. In June, WIRED reported face-recognition code in a companion app; the code was removed the next day. Meta said it was not pursuing the feature “right now.” Oslo’s text treats “ready to be integrated” as the live risk. Meta’s public line treats it as a plan that was paused. Readers can hold both documents and decide how much weight to give a pause.
A separate, older demonstration still sits in the background. In 2024 two Harvard students livestreamed from Ray-Ban Metas, piped the feed through a reverse-image tool and an language model, and retrieved names, ages and addresses of strangers on a campus and a subway. They did not need a Meta-built Name Tag. They needed a camera at eye level and software that already exists. That is the version of the problem that does not wait for a product launch. It is also the version that school rules can only partly touch.
What Mainstream Coverage Emphasised
The first global wave was an AFP dispatch. It quoted Midtgarden, listed the features of the glasses, noted the facial-recognition report, and explained the royal-truce delay. Outlets from Channels Television to GMA, SIC, ORF and BGNES ran that spine with little extra reporting. That is how a municipal education rule becomes a world brief: one official paragraph, one corporate name, one moral frame.
Norwegian broadsheets added the domestic politics. Aftenposten put Oslo and Bærum on the same page and reminded readers that Nordtun had already asked schools to act. NRK’s earlier file on Oslo’s 2025 school rules — no late arrivals without consequence, no photographing or filming other people without consent — shows the glasses ban sitting on top of a phone ban and a recording ban that the city already claimed as a win against bullying. Midtgarden has spent two years arguing that screens and covert clips make a school less safe. The glasses, in that telling, are the next object in a line, not a sudden conversion to Luddism.
Trade and tech desks in London and New York treated Oslo as one more pin on a map that already included Haberdashers’ Boys’ School, Woodbridge, Shrewsbury, U.S. districts and New York courts. Their frame is product risk and market size. Seven million units is a real business. A school ban in a capital of 700,000 people is not, by itself, an earnings event. A pattern of bans might be.
What Alternative and Campaign Coverage Emphasised
Outside the wires the nickname is uglier. British tabloids and a stretch of tech blogs have spent the summer on “pervert glasses”: clips of women and girls filmed without a glance of consent, then posted for sport. PetaPixel used that phrase in early September when it reported that Norway was considering a wider public ban. Futurism and school-safety writers documented teenage boys using the frames to record classmates. A Swedish designer, Mateusz Pozar, built a “Ban Ray” site after men in glasses approached and recorded his girlfriend. He sells stickers and signs and says the point is to force a conversation before the product becomes furniture.
That register is not the Oslo press release. Midtgarden did not use the word “pervert.” She used “extras in Meta’s data collection.” The campaign language and the municipal language still share a premise: the person in front of the lens did not agree to be a dataset. Where they split is motive. Campaign sites talk about sexualised filming and street harassment. Oslo talks about a business model. Both can be describing the same hardware. They are not describing the same harm with the same precision, and a serious reader should not pretend they are.
There is a third register that rarely makes the AFP paragraph. Accessibility groups note that the same glasses read rooms for blind and low-vision users. Meta has given pairs to blind veterans. A blanket school ban will catch a pupil who needs live description as well as a pupil who wants a covert camera. Oslo has not published a detailed exemptions policy in the material that reached the wires. Any honest account has to leave that as an open operational question, not as a solved footnote.
Children as Users, and Children as Subjects
The legal problem hiding under the politics is simple once it is named. Most child-privacy law was written for a child who opens an app. Smart glasses invert that. The child in the next desk never tapped “accept.” The International Association of Privacy Professionals put the point cleanly in a September back-to-school essay: the next wave of rules has to cover children as subjects of other people’s devices, not only as account holders.
That is not an abstract EU talking point. A German court has already said Meta’s generative models “inevitably capture” information about minors when adults post public photos of them. A U.S. class action over the glasses was amended this month to add “bystander” plaintiffs — people who never bought the product and say their faces, voices and bodies were taken anyway, including minors. Clarkson Law Firm, which is also behind earlier Meta privacy suits, described scenes the complaint says the camera can reach without anyone in the room having signed a terms-of-service screen. Those claims are allegations in a live case. They are not findings. They are part of the climate in which a Nordic city decided not to wait for the expert group.
Norway applies the GDPR through the EEA. Children’s data is not a free-for-all. Covert recording can already collide with criminal law, the Personal Data Act and school-environment duties. Oslo’s move is, in that sense, a conservative use of an existing tool: the school’s house rules. What it cannot do is rewrite the lawful basis for a camera walking past the school fence. That is why the vice mayor keeps pointing at the cabinet, and why Datatilsynet’s interest in a sales pause matters more than any single playground policy.
The same tension shows up wherever cameras leave the pocket and sit on the face. It is cousin to the argument in our earlier reporting on police radio-survey vans that sweep bystanders along with suspects, and to the fight over commercial satellites sold as daily remapping tools. Different hardware. Same question: who consented to be in the picture?
The Geopolitics Under a School Rule
Call this a playground story and you will miss why it belongs on a geopolitics desk. A Norwegian capital is telling an American platform that a consumer camera is not welcome around children. The platform is one of the largest advertising machines on earth. Its European headquarters sit in Ireland. Its glasses are cut in a Franco-Italian optics group. The data question is therefore not only “what does a thirteen-year-old film.” It is “which jurisdiction gets to decide what a face in public is worth.”
Europe has spent three years writing an AI Act and still cannot agree, in real time, whether a pair of sunglasses with a model attached is a prohibited social-scoring toy, a high-risk biometric system, or a gadget. National DPAs are told to enforce. Cities then do the thing states hesitate to do. That pattern will travel. If Oslo and Bærum hold, other Norwegian councils will copy the text. If Tung’s facial-recognition idea becomes a bill, importers will want a single EEA answer rather than 30 local ones. If they do not get it, the market will be a patchwork of school gates, court doors and cinema lobbies — the same patchwork already forming in the United States and Britain.
There is also a quieter industrial fact. Meta needs the glasses to become the next default screen. Phones matured. Headsets stayed niche. Eyewear is the bet that computing can live on a face without looking like a computer. A string of “not in this building” rules does not kill that bet. It does raise the cost of looking ordinary. Once a product needs a public-education billboard to explain its recording light, it has left the category of invisible tool. Oslo’s sentence about “extras” is an attempt to keep children out of that new default.
Readers who want the corporate side of the same firm in Europe can set this file next to our note on Zuckerberg’s Irish castle purchase near Meta’s European base. It is not evidence of anything in a schoolyard. It is a reminder that the company Oslo named is not an abstraction. It has staff, tax filings, product teams and a long argument with regulators about how long children’s data should live inside a model. A separate September analysis of a U.S. youth settlement made the same point in another jurisdiction: penalties and time limits can move, while the right to collect and target can stay intact. Oslo is trying to cut the collection step at the school door rather than bargain about it afterwards.
Arguments That Deserve to Be Heard Without a Team Shirt
The case for Oslo’s rule is straightforward. A school is a closed community of minors. Recording without consent was already a problem with phones; Oslo banned those too. Glasses make the same act harder to see. Parents cannot reasonably audit every pair of frames in a corridor. A bright-line ban is administrable. Teachers should not have to decide, in real time, whether a light is on. If facial recognition is even close to shipping, the moment to keep the hardware out of a classroom is before the feature, not after the first identified child.
The case against treating the ban as the last word is also straightforward. The devices help some disabled pupils and staff. A municipal rule does not bind a private school, a sports club or a public square. Criminal law already covers hidden cameras used for sexual humiliation; stacking a gadget ban on top of that can feel like punishing the object for the user. Meta can show a paper trail of LED enforcement and bricked cameras. Sales figures show millions of owners who are not filming locker rooms. Over-reading a school policy as a verdict on an entire platform is how a local tool becomes a culture war.
A third position sits between those two and is the one most cabinets actually occupy: regulate the feature, not the frame. Ban live facial recognition of strangers in public. Require a recording light that cannot be defeated. Treat school and courtroom as consent-free zones. Leave adult use on a hiking trail alone. That is roughly Tung’s public sketch. Midtgarden’s reply is that the feature and the frame are arriving together, and that “looking into the issue is not taking action.” Both sentences can be read without picking a party.
What should not be smuggled in is a claim the record does not support. There is no public Oslo study showing a measured spike in playground recordings with these specific glasses. There is no public Meta admission that schoolyard video is being ingested as training data from Norwegian children. There is a product that can record, a company that builds models on public imagery, a city that refuses to wait, and a government that prefers a committee. Those are the facts. The rest is inference. Inference is allowed. It should be labelled.
What Happens After a Ban That Already Started
Implementation will be dull, which is how school rules work. Staff will treat the glasses like phones. A first-week confiscation will become a local news photo. Parents of pupils who use the frames as a visual aid will ask for a written exception. Lawyers will ask whether a ban on “smart glasses” covers every brand or only camera glasses, and whether a teacher’s pair used for translation in a language class is in or out. Those fights are how a headline becomes a handbook.
The larger clock is national and European. The expert group will produce paper. Datatilsynet can open a file whether or not the cabinet likes the timing. The EDPB’s autumn report on social acceptability will give cover to whoever wants to move, or to whoever wants to wait. A British petition will not bind Oslo, but it will keep the nickname “pervert glasses” in search results next to Midtgarden’s “extras.” Meta will keep shipping LED patches and awareness ads. Google and Snap will keep preparing their own pairs. None of that waits for a Norwegian white paper.
For a city that already took phones out of primary and lower-secondary days, the glasses were the next object that looked like a face and acted like a logger. For a company that needs the next surface after the smartphone, a capital city calling its customers unpaid extras is a branding problem that no capture LED quite solves. For a parent standing at a fence, the useful question is smaller than either speech: can my child spend six hours in a yard without becoming a clip? Oslo answered no, not with these frames. The cabinet has not yet answered in a statute. That is the whole present tense of the story.
Sources
- Primary Norwegian reporting: Aftenposten, 10 September 2026
- International wire: AFP via Channels Television, 10 September 2026
- Latest Europe wrap: AFP / Geo.tv, 12 September 2026
- Education minister’s earlier position: TV 2, 12 August 2026
- Digitalisation minister on facial recognition: VG, 25 August 2026
- Meta LED enforcement: Semafor, 1 September 2026
- Facial-recognition reporting: New York Times, 13 February 2026
- Civil-society warning: WIRED, 13 April 2026
- Related Planet Today files: FalcoNet surveillance vans, Merlin satellites, living-neuron data centre, Meta in Ireland, AI risk debate
Original source
Core quotations and the 27 August / 10 September timeline follow the AFP dispatch published 10 September 2026 and the same-day Norwegian account in Aftenposten, cross-checked against Bærum’s school-equipment notice and prior TV 2 / VG interviews with national ministers. King Harald V’s death on 28 August 2026 and the state funeral on 9 September 2026 are on the public royal and Reuters record.
Note for fact-checkers
AFP is a wire service. Aftenposten, VG, TV 2 and NRK are mainstream Norwegian outlets with ordinary editorial standards and ordinary political leanings; they are not underground samizdat and they are not a government bulletin. Meta’s LED statements come from named executives in Semafor, The Verge and CBS. Facial-recognition plans were reported by the New York Times from unnamed people and later bounded by Meta’s public “not right now.” Campaign sites and tabloids that use “pervert glasses” are describing a real pattern of non-consensual clips; they are also running a moral brand. Treat each layer as what it is. European state media and Russian state media both editorialize. So do commercial platforms. The check is always the same: can the quote be found, can the date be found, and is the inference marked as inference.
Original article: Oslo Bans Smart Glasses in Schools Over Meta Data Collection on Planet Today 🚀
Automatically republished from the main blog.