"It's not always obvious to the consumer what it is they're giving up in terms of privacy to enjoy these new, efficient consumer products that make their lives more productive and more frictionless," said Ryan Clarkson, whose firm represents consumers suing Meta and Luxottica of America over allegedly misleading privacy claims involving their smart glasses.
The lawsuit alleges that human annotators viewed and labeled recordings of people changing clothes, using bathrooms and handling financial information after footage captured by the glasses was uploaded to Meta's servers to train its AI systems.
Those recorded often appeared unaware they were being filmed, Clarkson said. "No reasonable consumer who was promised privacy could ever have concluded these devices would intentionally or inadvertently capture them in their most private moments," he told Reuters.
Luxottica and OpenAI did not immediately respond to requests by Reuters for comment.
Meta said a light that cannot be turned off during filming protects the privacy of people around the user, and that the camera is disabled if the light is covered or tampered with.
The company also said it disagrees with the allegations in the lawsuit and will contest them, adding that reviewing user data to improve products was a common practice across the industry and that Meta filters out identifying information.
The privacy backlash has led other companies to consider glasses without cameras.
Even Realities, a Shenzhen-based startup founded by former Apple engineer Will Wang, who worked on the Apple Watch, introduced its camera-free G1 glasses in 2024. Its latest flagship, the G2, launched in November 2025.
Wang told Reuters in July that consumers increasingly associate smart glasses with covert recording and that cameras should wait until companies can put stronger safeguards in place. The company, whose valuation hit $1 billion in July, expects to end the year with a few hundred thousand users. Consumers account for about 90% of its sales, Wang said.
San Francisco-based Pocket is betting that a phone-mounted recorder that requires users to press a button to capture conversations can ease privacy concerns around audio-only devices by making recording as deliberate and visible as using a phone camera. The company said that about 170,000 Pocket devices were in use eight months after launch, including by workers at firms such as DoorDash, Michelin and Santander.
Plaud, which sells card-shaped and wearable AI recorders, has also prioritized privacy with Plaud One, a $249.99 Explorer Edition earbud-and-case system that records only when activated and alerts bystanders through lights or an audio announcement. The company has 2.5 million users and more than $100 million in annual recurring revenue.
Privacy concerns sparked by AI devices are not unprecedented, said Jitesh Ubrani, director at research firm IDC, noting that smartphones faced similar unease when cameras and microphones first became ubiquitous.
"As a society, we want to make sure there’s consent and awareness when people are being recorded," Ubrani said. "That’s different from being opposed to recording altogether."
Privacy fears are slowing the AI-device push
Companies betting AI devices will usher in a post-smartphone era are fueling privacy fears around the globe that could slow adoption of the emerging technology.
The backlash against devices that can see, hear and respond to the world around them has been strongest against Meta's Ray-Bans, which made smart glasses mainstream by pairing EssilorLuxottica's trendy designs with AI from the social media giant chasing "personal superintelligence."
The glasses have become a lightning rod for growing unease over AI-powered devices, drawing scrutiny from courts weighing privacy lawsuits, users fearing covert recordings and critics warning of a future in which ubiquitous AI eyewear erodes expectations of anonymity and consent. Several UK venues, including one of the country's biggest pub chains, Wetherspoon, have restricted the use of the frames.
Meta is expected to unveil camera-free smart glasses as early as Wednesday at its Connect conference, according to media reports, highlighting the balancing act facing companies seeking a share of the more than $600 billion smartphone market.
Device makers are trying out a range of form factors to get past those concerns, from glasses and pins to pendants and voice recorders that clip magnetically to a phone, according to Reuters interviews with nearly a dozen technology executives, analysts and legal experts. Some firms are finding faster traction with businesses than consumers.
The outcome of the privacy debate could shape a market that tech firms see as the next major computing platform. If companies fail to win public trust, the same features that make AI-powered wearables useful — always-on cameras, microphones and environmental awareness — risk becoming the barrier to their adoption.
Big tech and startups are adjusting their product plans
"The winning form factors will be the ones that make interaction more natural, useful and trusted," said Jay Kim, executive vice president at Samsung's mobile division, in an interview with Reuters in July. As AI takes on more tasks, he said, users will demand systems that remain secure and under their control.
Samsung plans to enter the smart-glasses market later this year through partnerships with Gentle Monster and Warby Parker, but does not view the devices as replacements for smartphones.
For some startups, however, solving the privacy problem may be existential.
Speaking with journalists in December, OpenAI CEO Sam Altman described his vision for an always-on device capable of understanding a user's surroundings while knowing when to help and when to stay silent. He argued that such products would ultimately need to process data locally rather than continuously send information to the cloud.
"You want a device that you can trust ... that can have full context and know when to be helpful, when not to," he said. "If a device is going to be listening to your whole life and watching your whole life, most people will not want that running on a cloud server." He added that local AI processing "may take a couple of years longer" depending on how quickly OpenAI can scale its processors.
Experts say the central challenge is that AI devices become more useful as they collect more information about their surroundings. That information can include footage of people who never consented to being recorded.
META•
