❌

Normal view

There are new articles available, click to refresh the page.
Yesterday — 25 September 2026TechRadar - All the latest technology news

Smart glasses were being used to secretly film women 12 years ago

One of the advantages of having written about tech for as long as I have is that you start to notice when the same ideas come around decade after decade, in different shapes and sizes. Especially the bad ones.

12 years ago Google invented something called Google Glass — a set of glasses complete with a camera so you could record what you were looking at — and it was so roundly hated by everyone that we all breathed a collective sigh of relief when they went away.

But now it's 2026 and smart glasses are back, powered by AI, and tied in with a world-famous fashion brand, but fundamentally they still have exactly the same problem, which still hasn't been fixed: they enable people to record other people without their consent.

The problem is that none of this should have come as a surprise. The privacy risks of putting a camera on somebody's face weren't merely predictable — they were predicted. Women were raising concerns about men using Google Glass to record them more than a decade ago.

And yet here we are again.

The curious case of Google Glass

Google Glass

(Image credit: Shutterstock / Hattanas )

Let’s remind ourselves what happened the first time that smart glasses arrived.

Google Glass arrived in 2013. It puts a small heads-up display and camera on the wearer's face, allowing them to take photos and record video from their point of view. Almost immediately, privacy became one of the defining controversies around it.

By 2014, women were explicitly raising issues of abuse. Wired reported that women were particularly interested in technology that could detect and block Glass because of concerns about men covertly recording them in nightclubs. "Even if they didn’t know if the device was recording, they felt threatened by its presence", said Oliver, the creator of the detection device.

Google Glass wearers were being called “Glassholes,” businesses were banning the device, and Google itself was publishing etiquette guidance because of complaints about people creepily filming others without permission.

By 2015, consumer Glass effectively died. Privacy wasn't its only problem — its price, awkward appearance, and lack of compelling everyday uses all played a part — but by January 2015 Google had stopped selling the consumer Explorer Edition. Glass survived as an enterprise product before Google finally discontinued the Enterprise Edition in 2023.

Smart glasses didn’t die with Google Glass. They had simply retreated to a lab to lick their wounds and wait for that vital next piece of technology that would make them relevant again. That technology was AI. In 2023, AI-powered smart glasses, partnered with leading fashion brands, started to emerge, with Meta leading the charge with its Ray-Ban Meta glasses.

This time they looked much more like ordinary glasses. Meta had also found the compelling use cases that Google Glass struggled to provide: its glasses could use AI to answer questions about what you were looking at, give you directions, and translate signs and conversations. But they could also still record what, or more importantly, who, you were looking at without needing that person's consent.

It’s like women don’t exist

Meta Ray Ban glasses being worn.

(Image credit: Shutterstock / Alexander Fedosov)

Frankly, I don’t think Meta should have released them with that ability, but here we are. It’s like we’ve learned nothing from the Google Glass debacle. Almost as soon as the next generation of smart glasses emerged they were being used to harass women and record them without their consent or knowledge.

Researchers at the University of Sydney have been studying hundreds of social-media videos involving smart glasses and found men using them to approach and record women without their knowledge, including at gyms, workplaces and beaches. Australia's eSafety Commissioner says the technology can amplify gendered harms because recording can be almost undetectable, and specifically highlights pickup-artist content in which women often don't know they're being filmed.

The researchers analysed 350 publicly available Instagram videos from 2023–2026. In a subset of covert POV videos, around 60% of interactions were classified as potential harassment. And in about 43% of the videos, women were subjected to further abusive online commentary, sometimes exposing personal information.

These are real victims. CBS interviewed a woman who discovered she'd been secretly recorded only after the resulting video went viral, attracting more than 200,000 views. Her reaction said it all: “I had no say.”

And France has now opened an investigation involving alleged sexual harassment and discreet smart-glasses recording.

I don't think these products were designed to harm women. That's almost the point. Too often, the technology industry seems to design products around what the person using them can do, without giving enough consideration to the person standing on the other side of the technology. And when the potential for abuse disproportionately affects women, that's a particularly serious blind spot. Women have historically been underrepresented in the data and testing used for product design, with designers still describing the male body as the implicit default.

Meta clearly did think about privacy. The glasses have a recording LED specifically designed to tell people nearby when the camera is being used. But if your safeguard against covert recording is a tiny light on the front of the glasses — one that people have found ways to conceal or disable — is that really enough?

Did it consider what happens when a man wearing an almost invisible camera approaches a woman who's alone?

Meta's audio-only smart glasses

The Ray-Ban Meta Audio glasses

Meta Ray-Ban Audio smart glasses. (Image credit: Meta)

Yesterday at Meta Connect, Meta announced a new version of its smart glasses called the Ray-Ban Meta Audio glasses, which don’t have a camera. They got barely a mention in Zuckerberg’s keynote, but I wonder if they were released because of the existential threat to its business that the now commonly used “perv glasses” moniker poses.

Ray-Ban Meta Audio has no camera. You can still listen to music, make calls, translate conversations, and talk to Meta's AI, but the person sitting opposite you no longer needs to wonder whether they're being secretly recorded.

Meta hasn't said the Audio glasses exist because of the privacy backlash, and removing the camera inevitably removes some genuinely useful features too. For visually impaired people in particular, a camera combined with AI can be transformative.

But Ray-Ban Meta Audio proves something important: smart glasses don't have to contain a camera.

Twelve years ago, women were already warning that putting inconspicuous cameras on people's faces could enable men to record them without their knowledge, and today researchers are documenting precisely that behavior.

Google Glass users became known as “Glassholes” because people hated that the person they were talking to was recording them. Twelve years later, we're having the same argument about “perv glasses.”

That's what makes me angry. This wasn't an unforeseeable consequence of some revolutionary new technology. We'd already run this experiment and realized that these should never have got out of the lab in the first place.

Before yesterdayTechRadar - All the latest technology news

ChatGPT got GPT-6, but you can't use it in Chat — confused?

ChatGPT just got GPT-6, except there’s a catch: you can’t actually use it in Chat.

If that sentence makes absolutely no sense to you, I don’t blame you. Until now, you could use ChatGPT quite happily without giving much thought to the distinction between Chat and Work functionalities. I certainly did. It was just that little slider at the top of the screen I never bothered with.

I suspect most of you are the same. Chat was where I talked to ChatGPT, while Work was something sitting alongside it that I could largely ignore because the Chat setting let me do almost everything I wanted.

GPT-6 changes that. OpenAI’s newest model has arrived in ChatGPT, but instead of appearing in the familiar model picker on the right hand side of the prompt bar, GPT-6 Astra, Sol and Luna are available in Work and Codex only. So, naturally, the first thing I did was go looking for them.

And somewhere between opening Work, figuring out what OpenAI actually expects me to do there, and finally getting my hands on GPT-6, I realized this launch is about more than a new model. OpenAI has just made the difference between chatting with ChatGPT and asking it to work for you in a way it never really did before.

When to Chat and when to Work

The easiest way I’ve found to understand the difference is to think about how much of the job I want ChatGPT to take responsibility for. In a normal Chat conversation, I’m usually working alongside the AI. I ask something, look at the response, add more information and gradually steer it towards what I need.

Work feels different because I’m handing over more of the process. I can give ChatGPT a larger task involving multiple steps and let it work out how to approach it, which might mean researching information, looking through files or using websites rather than waiting for me to tell it what to do at every stage.

If I’ve got a bigger job involving several files or lots of moving parts, I’ll often start there and let ChatGPT work out what it needs before I get involved again. Chat feels collaborative; Work feels much more like delegation.

I wouldn’t nominate OpenAI for any interface design awards because one of the reasons that the distinction between Work and Chat is so confusing is that they both use exactly the same interface, with a prompt bar at the bottom of the screen.

In fact, you can choose Work mode and then just start chatting to the AI as you normally would, although I've tried this, and it's not a particularly good experience. Responses can take considerably longer than they do in Chat, because Work is designed around longer, multi-step jobs rather than firing back quick conversational answers. If I just want to ask ChatGPT something, Chat remains the obvious place to do it.

And once you understand that distinction, putting GPT-6 Astra, Sol and Luna in Work rather than Chat starts to look considerably less strange.

ChatGPT on a mobile phone.

(Image credit: Apple / OpenAI)

So what exactly is GPT-6?

OpenAI calls GPT-6 Astra “the most intelligent and aligned model in the world”. Think of it as the heavyweight of the family: OpenAI's most capable model, built for the really difficult jobs where you want maximum intelligence rather than maximum speed. OpenAI particularly emphasizes Astra's ability to operate software, browse, conduct research and complete multistep professional workflows.

And because not every job requires the power of a burning sun to complete, OpenAI has also released two smaller versions called GPT-6 Sol and GPT-6 Luna. These models are more lightweight and less expensive to use. They’re trained using similar methods to GPT-6 Astra, but OpenAI say they “build on the advances behind GPT-6 Astra” and bring much of its strengths into faster and more affordable models.

There is one wrinkle here. GPT-6 Astra has actually been around since earlier this month, and GPT-6 Pro, which is powered by Astra, is available in regular Chat on some higher-tier plans. I'm a ChatGPT Plus subscriber, however, which means my access to Astra — and now the new GPT-6 Sol and Luna models — is through Work and Codex.

If I switch to the Work tab in my Plus account then in the model picker I get access to GPT-6 Luna High, GPT-6 Sol Light, GPT-6 Sol Medium, GPT-6 Astra Light and GPT-6 Astra Medium.

Until now, I could happily spend all my time in Chat, using GPT-5.6 Sol, and largely ignore Work. GPT-6 changes that. OpenAI isn't just giving us smarter models; it's starting to separate the AI we talk to from the AI we give jobs to. If its most powerful new models are going to live on the Work side of that divide, then that little switch I've spent months ignoring suddenly matters a lot more.

I finally got Siri AI on my iPhone, so I gave it 5 jobs I normally use ChatGPT Voice for — I wasn’t expecting it to be this good

A few days ago, I wrote about how upgrading to iOS 27 doesn’t automatically give you the new Siri AI. To get Apple’s upgraded assistant, you have to jump through an additional hoop: go into Settings, find Siri and apply to join the waitlist for the Siri AI beta.

The good news is that I didn’t have to wait long. Yesterday, my application was approved and Siri AI quietly installed itself on my iPhone.

It feels like years since Apple first talked about giving Siri a much-needed intelligence upgrade, and I almost can’t believe it’s finally here. Well, almost here — it remains a beta — but the important thing is that it does what it says on the tin.

Before Siri AI, “Would you like me to use ChatGPT to answer that?” had become Siri’s standard response to almost anything interesting I asked it to do beyond setting a timer. Now Siri can produce proper answers of its own.

You can still activate Siri AI in the familiar way by saying “Hey Siri,” but there’s also a new Siri app on the Home Screen. Inside the app, you can either type to Siri or talk using your voice, much as you can with ChatGPT.

Things look different, too. A new floating globe animation appears when Siri AI is listening. There are so many things it should now be capable of that I almost didn’t know where to start.

So I kept it simple. I picked five things I regularly use ChatGPT Voice for and tested whether Siri AI could now handle them.

This isn’t a comparison of everything ChatGPT can do. Outside Voice mode, ChatGPT is capable of far more. I wanted to compare the two assistants as I would actually use them on my iPhone: by talking to them.

Siri AI is still technically in beta, but Apple has decided it’s ready enough for people to use, so I’m going to test it.

Test 1: Can it actually have a conversation?

Siri AI on an iPhone

(Image credit: Apple)

The first test involved finding information and, more importantly, explaining it clearly. I asked Siri AI about something I’d written about a few days earlier: AI voice-cloning scams.

“Why are people worried about AI voice-cloning scams, and what should I actually do to protect myself?” I asked.

In response, Siri AI produced a massive slab of text about voice scams, displayed it on screen and read it aloud in a slightly stilted voice. The information was accurate and factual, but the response was long and boring, and Siri’s voice had no real personality.

I interrupted it with my next question: “What is the single most important thing I should actually do?” To its credit, it stopped and answered the new question.

ChatGPT Voice, by comparison, talked to me like a human who was interested in the subject. It sounded conversational rather than as though it was reading from a block of text. It didn’t waffle and gave me the information I needed neatly and succinctly.

Verdict: Siri AI can now deliver useful information on its own, but it still doesn’t feel like a proper conversation, and its voice needs work before it sounds natural.

Score: 3/5

Test 2: Does Siri know what’s happening in my life?

Siri AI searching Mail.

(Image credit: Apple)

For the second test, I asked Siri to retrieve something I vaguely remembered from my Gmail inbox, which I access through Apple’s Mail app, rather than giving it precise keywords.

“Find the email with my tickets for Hubris in October and tell me what date it’s on.”

Siri AI checked Mail but initially couldn’t find anything, so I followed up with: “They’re at the Griffin.” Sure enough, it then found the tickets in my inbox and gave me the information I needed.

Next, I asked: “What do I need to bring?”

Siri AI told me I would need to have my tickets downloaded to my phone or printed out. Once again, its answer was functional and cold — and focused entirely on the tickets. There was no consideration of anything else I might need for a concert.

I fired up ChatGPT Voice to compare the answers, and this is where Siri AI had a major advantage. While text-based ChatGPT can connect to my Gmail and search my inbox, ChatGPT Voice can’t currently do that. It therefore failed to retrieve the information I wanted, although it performed a web search and found the date of the concert.

ChatGPT shone when I asked, “What do I need to bring?” Its answer was like night and day compared with Siri’s because it offered genuinely useful advice:

“For a night like that, comfy clothes plus good shoes are key. Maybe a light jacket. Bring the ticket on your phone with a backup screenshot. Photo ID, card for the bar and maybe earplugs. And if you’re meeting anyone, set a rough time and place outside, just in case.”

Verdict: Siri AI beat ChatGPT Voice at the practical part of the test because it could search the Gmail account connected to Apple Mail. However, ChatGPT was much better at understanding the wider context and anticipating what I might actually need for the concert.

Score: 3/5

Test 3: Can Siri see what I’m seeing?

Siri AI is supposed to be able to “see” what’s on your screen and answer questions about it. This is potentially one of its most useful new features, so I gave it a thorough test.

I opened a range of things on my screen—a document, an email, a graph, a photo and a webpage—and asked: “What’s the important thing I need to know here?” Each time, Siri AI successfully picked out the relevant information.

While browsing Crunchyroll, the popular anime streaming service, it gave me a decent explanation of each anime featured on the screen. Siri talks to you while simultaneously displaying its answer over whatever you’re viewing. For example, this is how it looked when describing the anime Black Torch, which happened to be featured on screen:

Siri AI describing what it sees in the Crunchyroll app.

(Image credit: Apple / Crunchyroll)

When I opened an Amazon product page for an unavailable item, Siri AI correctly identified that the most important detail was that it was out of stock. When I viewed an invoice from my accountant, it correctly identified the amount and the payment deadline.

Siri AI could also see through the camera and identify objects very well. Being able to invoke it with “Hey Siri”—or simply move the slider from Photo to Siri while the Camera app is open—is genuinely convenient. It involves much less messing around than entering Live mode in Gemini or opening the camera through ChatGPT Voice.

Verdict: I’ve got to say, Siri AI smashed this test. I’m going to use this feature all the time.

Score: 5/5

Test 4: Can it think with you?

Siri AI creating a workout.

(Image credit: Apple)

Next, I gave Siri an open-ended problem that I would normally bring to ChatGPT:

“I’ve got 30 minutes tonight, don’t want to go to the gym, but want some exercise that’ll actually raise my heart rate. What should I do?”

I then changed the rules:

“Actually, I’ve only got 15 minutes and I don’t want to leave the house.”

This is the kind of test that reveals whether Siri has become useful for open-ended judgment rather than simply following commands.

Siri recommended a 30-minute high-intensity interval training workout. It presented me with a simple plan and then asked whether I would like to start a timer so I could begin. When I said I had only 15 minutes, it adapted the workout accordingly. Not bad.

ChatGPT Voice took a different approach. It didn’t give me a workout plan in advance; instead, it immediately launched into a session that it wanted me to follow along with.

That was quick and direct, but I preferred Siri AI’s approach of showing me the workout before asking me to begin. Siri can combine text and voice in a way that ChatGPT currently doesn’t in Voice mode, and that worked particularly well here.

Verdict: Siri AI and ChatGPT Voice approached the task differently, but I preferred Siri’s ability to present the plan visually before helping me get started.

Score: 4/5

Test 5: Can it actually do something ChatGPT can’t?

Searching Mail with Siri AI.

(Image credit: Apple)

One home-field advantage Siri AI should have is the ability to perform multistage tasks involving real iPhone apps, so I asked it:

“Find the details of the next concert I have tickets for, add it to my calendar and remind me the evening before.”

Siri AI worked like a dream:

“The next concert you have tickets for is The HU on Thursday, October 1, 2026, at 19:00. I’ve added this to your calendar and set a reminder for the evening before.”

The HU are a Mongolian folk-metal band — so no, Siri hadn’t misheard The Who — and that was a perfect execution of the task.

ChatGPT Voice couldn’t search my Gmail, so it had to rely on previous conversations and web searches. It got close, but it needed more details from me to complete the request. It also told me: “I can’t add it to your iPhone calendar from here.”

Verdict: Siri AI’s system-level integration allowed it to behave like a true virtual assistant, completing a useful multistage task that ChatGPT Voice simply couldn’t perform.

Score: 5/5

Will I keep using Siri AI?

What became clear during these tests is that Siri AI’s greatest strength is its position inside the iPhone.

It can see what I’m doing, access information held inside my apps and turn a spoken request into an action without making me copy details between different services. That gives Siri an immediate advantage over ChatGPT Voice on iOS.

Siri AI still has a long way to go before it sounds as natural as ChatGPT Voice. ChatGPT remains the assistant I would choose when I want to explore an idea, work through a problem or simply have a convincing conversation. Outside Voice mode, ChatGPT is also much more capable and can even complete bookings or code apps to solve problems.

But when I want something done on my iPhone, Siri is suddenly the more useful option.

Apple has done enough to make me start using Siri again, and after years of disappointment, that might be the most surprising result of all.

❌
❌