AI wearables that record conversations and capture video are becoming mainstream, yet the privacy concerns they raise may be impossible to fully resolve. Major tech companies including Meta, Amazon, and Samsung are betting billions on glasses, bracelets, and clips that can see and hear your world in real time, feeding that data to AI assistants. A CNN journalist who tested three leading devices over several months found them genuinely useful in narrow situations, but discovered that the social friction and privacy risks may outweigh the benefits for most users.

What Are These AI Wearables Actually Designed to Do?

The vision from tech leaders is straightforward: wearable devices with cameras and microphones give AI assistants continuous access to your surroundings, allowing them to offer real-time advice, identify landmarks, transcribe conversations, and eventually complete tasks on your behalf. Meta is already selling millions of its Ray-Ban glasses with embedded cameras and microphones through EssilorLuxottica. Samsung plans to launch its own smart glasses later in 2026. Amazon's Bee Pioneer is a wristband that records conversations and offers suggestions. Plaud's Notepin S is a small clip-on recorder designed to capture meetings and social interactions.

The appeal is clear: imagine walking past a historic monument and instantly learning its history without pulling out your phone, or attending a networking event where your wristband automatically suggests which contacts to follow up with afterward. These scenarios played out during testing. But the friction came from something less tangible: the discomfort of asking friends, coworkers, and strangers for permission to record them, and the creeping sense that the devices were learning things about you that you never explicitly shared.

Why Are Privacy Advocates So Concerned About These Devices?

Unlike smartphones, which are obvious when in use, wearable cameras and microphones are difficult to spot. A person wearing glasses or a bracelet doesn't signal to others that they're being recorded. This fundamental asymmetry breaks the traditional consent model that has governed recording and photography for decades.

"So the whole notion of consent is kind of disintegrating," said Irina Raicu, director of the internet ethics program at Santa Clara University's Marukkula Center for Applied Ethics.

Irina Raicu, Director of Internet Ethics Program, Santa Clara University's Marakkula Center for Applied Ethics

The risks extend beyond awkward social moments. Women have reported that men have used Meta's AI glasses to film and post videos of them on social media without consent. Wearable devices with microphones and cameras can be worn in private spaces like dressing rooms, creating opportunities for serious misuse. While Meta, Amazon, and Plaud have all added indicator lights to show when recording is happening, these lights are often difficult to see because most people aren't looking at your wrist or shirt during a conversation.

"There are some real concerns with how this could be misused in dangerous ways," said Calli Schroeder, senior counsel and director of the AI and Human Rights program at the Electronic Privacy and Information Center.

Calli Schroeder, Senior Counsel and Director of AI and Human Rights Program, Electronic Privacy and Information Center

How Are AI Wearables Learning Things You Never Told Them?

One of the most unsettling discoveries during testing involved the Amazon Bee bracelet's ability to infer emotional states and personal concerns from overheard conversations. During a casual board game night, the journalist mentioned moving to a new apartment in the fall. Days later, the Bee app reminded her to book movers for October, which was helpful. But then Bee suggested that her words indicated anxiety about "stability" and her "sense of home," recommending she photograph "one beloved corner" of her current space to take with her. The problem: she wasn't anxious about the move at all and certainly hadn't intended to convey that to anyone, let alone an AI bracelet.

This scenario illustrates a deeper issue with always-listening AI assistants. They're designed to surface insights based on what they learn about you over time, but early on, when they have limited context, those insights can be wildly inaccurate. Amazon acknowledged this in a statement to CNN, noting that the Bee experience will improve as the system learns more about individual users. But the underlying problem remains: these devices are making psychological inferences about you based on fragments of conversation, without your explicit consent or awareness.

Steps to Protect Your Privacy When Using AI Wearables

• Check Recording Indicators: Before using any wearable with a camera or microphone, verify that indicator lights are visible and functioning. Meta's Ray-Ban glasses have a bright white LED that blinks when recording, and it has no off switch, but lights on wristbands and clips may be harder to spot in normal lighting.
• Ask for Explicit Permission: Even if a device has an indicator light, ask people directly before recording conversations or taking photos of them. This is especially important in professional or social settings where recording norms are still evolving.
• Review Device Settings Regularly: Check what data your wearable is collecting and how it's being used. Mark inaccurate insights as wrong to help the AI learn your preferences, and disable features you don't need.
• Understand Data Sharing Policies: Read the privacy policy for any wearable device to understand whether your recordings are stored on-device, sent to company servers, or shared with third parties for training AI models.

What Do Tech Leaders Say About the Future of Wearable AI?

Despite privacy concerns, major tech companies are doubling down on wearable AI. Qualcomm CEO Cristiano Amon told CNN in June that some of the largest companies in the world are developing gadgets like AI pendants, pins, and jewelry. For Amazon, the Bee bracelet is central to the future of its Alexa assistant. The company's reasoning is that an AI assistant with context about your entire day, not just time spent at home, can complete tasks more effectively and help you do what's next.

"The more they know, the better off they can complete tasks for you, and help you do what's next. That's quite powerful," said Panos Panay, Amazon's senior vice president of devices and Alexa.

Panos Panay, Senior Vice President of Devices and Alexa, Amazon

OpenAI is also working on a family of hardware gadgets, according to President Greg Brockman. He suggested that as AI agents advance, the days of keyboards and mice will soon be behind us. Plaud's CEO Nathan Xu believes that recording devices could eventually help create a shorter work week by giving AI models better understanding of jobs with less prompting, potentially helping people work faster.

Meta CEO Mark Zuckerberg has framed AI glasses as a cognitive tool. He suggested on an earnings call that people without AI glasses or some way to interact with AI may be at a "significant cognitive disadvantage" compared to those who do. This framing positions wearable AI not as optional convenience, but as a competitive necessity.

Why the Social Friction Matters More Than You Might Think

The journalist who tested these devices found them genuinely helpful in specific scenarios. Meta's Ray-Ban glasses provided instant information about landmarks during a trip to Lisbon. The Amazon Bee bracelet was useful at networking events, suggesting contacts to follow up with. Plaud's Notepin S made it easier to record meetings without holding a phone. Yet all three devices spent significant time buried in backpacks or desk drawers because using them socially felt strange.

This friction is not a minor inconvenience. It suggests that even when privacy concerns are addressed with indicator lights and consent mechanisms, there's a deeper social discomfort with always-recording devices that tech companies may struggle to overcome. The gap between what tech leaders envision and what consumers actually want to wear remains substantial. Until that gap narrows, wearable AI may remain a niche tool for specific use cases rather than the mainstream computing platform that companies like Meta and Amazon are betting on.