Meta wants to put cameras on every face in America. It just doesn’t want anyone asking uncomfortable questions about what those cameras can do.
In a move that has alarmed privacy advocates and civil liberties organizations, Meta has been quietly rebuffing requests from more than a dozen civil society groups to implement meaningful safeguards against facial recognition abuse on its Ray-Ban Meta smart glasses. The groups — which include the ACLU, the Electronic Frontier Foundation, and a coalition of digital rights organizations — have spent months pressing the company for commitments on privacy protections. Meta’s response, delivered in a letter this spring, amounted to a polite but firm refusal to engage with their most pressing concerns.
The tension sits at the center of one of the most consequential technology debates of the decade: as AI-powered wearable cameras become mainstream consumer products, who bears responsibility for preventing their misuse? And can a company that has spent billions building the world’s most sophisticated facial recognition infrastructure credibly promise it won’t be used?
As Wired reported, the coalition sent a detailed letter to Meta in January 2025, laying out specific requests including a prohibition on facial recognition features, transparency about what visual data gets processed by Meta’s AI systems, and a commitment not to use images captured by the glasses to train AI models. The groups wanted concrete policy commitments, not vague assurances. What they got back was something closer to a corporate shrug.
Meta’s response, according to Wired’s reporting, acknowledged the groups’ concerns but declined to make binding commitments on most of the key asks. The company pointed to existing privacy features — a small LED light that illuminates when the glasses are recording, and the ability for users to control their own data — as sufficient protections. Privacy advocates found this deeply inadequate. The LED indicator, they argue, is nearly invisible in daylight, easy to cover, and meaningless to people who don’t know what it signifies.
“A tiny light on a pair of sunglasses is not informed consent,” one advocate involved in the discussions told reporters.
The timing of Meta’s stonewalling matters. The company is aggressively expanding its smart glasses lineup. At its Connect conference last year, Mark Zuckerberg made clear that wearable AI devices represent Meta’s next major hardware bet, potentially even more important than its virtual reality headsets. The Ray-Ban Meta glasses, which retail for around $299, have exceeded internal sales expectations. And Meta is reportedly developing a partnership with Oakley to produce sport-focused smart glasses, broadening the product line’s appeal to athletes and outdoor enthusiasts.
More models. More cameras. More faces captured without knowledge or consent.
The civil society groups aren’t operating on abstract fears. In October 2024, two Harvard students made international headlines when they demonstrated a system called I-XRAY that paired Meta’s Ray-Ban smart glasses with the facial recognition search engine PimEyes to identify strangers in real time. Walk up to someone on the street, glance at them through the glasses, and within seconds pull up their name, address, phone number, and social media profiles. The students built it as a proof of concept to illustrate the dangers. It worked disturbingly well.
That demonstration crystallized what privacy researchers had been warning about for years. The hardware is already in consumers’ hands. The facial recognition databases already exist. The only thing standing between the current moment and ubiquitous real-time identification of strangers is software integration — and that barrier is getting thinner by the month.
Meta has publicly stated it has no plans to integrate facial recognition into the Ray-Ban Meta glasses. But the company’s history with the technology makes that assurance ring hollow. Meta built and deployed one of the largest facial recognition systems ever created through Facebook’s tag-suggestion feature, which automatically identified people in uploaded photos. The system was so effective — and so invasive — that it became the subject of a landmark class-action lawsuit in Illinois under the state’s Biometric Information Privacy Act. Meta settled that case in 2022 for $650 million.
The company subsequently announced it was shutting down its facial recognition system and deleting the faceprints of more than one billion users. But shutting down a product isn’t the same as abandoning the underlying capability. Meta retains enormous expertise in computer vision and facial analysis. Its AI research division continues to publish papers on visual recognition systems. The infrastructure could be rebuilt.
And third parties don’t need Meta’s permission to build on top of its hardware. That’s precisely what the I-XRAY demonstration showed. As long as the glasses can capture and transmit images — which is their core function — external developers can pair them with any facial recognition service available on the internet. PimEyes, Clearview AI, and dozens of smaller companies maintain massive databases of scraped facial images. Meta doesn’t need to build facial recognition into its glasses for the glasses to become facial recognition devices.
This is the argument the civil society coalition has been making, and it’s the argument Meta seems unwilling to seriously address. The groups have asked for technical measures that would make it harder to use the glasses for unauthorized identification — things like on-device restrictions on how captured images can be shared with third-party applications, or watermarking systems that would allow individuals to detect when they’ve been photographed by smart glasses. Meta has shown little interest in these proposals.
The company’s reluctance has a clear business logic. Every restriction on what the glasses can do makes them less useful as a product. Meta’s AI assistant, which is integrated into the glasses, relies on the camera to answer visual questions — “What kind of flower is this?” or “Can you read that sign for me?” Limiting how images flow through the system could degrade these features. And Meta is betting that multimodal AI — systems that can see, hear, and respond to the physical world — will be the defining consumer technology of the next decade. Hobbling the camera would mean hobbling the strategy.
But the privacy implications extend well beyond individual consumer choice. Smart glasses create what surveillance scholars call an “asymmetric visibility” problem. The person wearing the glasses knows they’re recording. Everyone around them does not. Traditional cameras — security cameras, phone cameras — are generally visible and understood. A pair of Ray-Bans that look nearly identical to non-smart Ray-Bans creates a fundamentally different dynamic. You can’t opt out of being recorded if you can’t tell recording is happening.
This asymmetry has already caused friction in social settings. Reports from early adopters describe being asked to remove the glasses in bars, restaurants, and private gatherings. Some venues have begun posting policies prohibiting smart glasses. But these are ad hoc responses to a systemic problem, and they don’t scale.
The regulatory picture in the United States remains fragmented. Illinois’s BIPA remains the strongest state-level biometric privacy law, but it applies primarily to the collection and storage of biometric identifiers by private entities, not to the act of photographing someone in public. A handful of cities — San Francisco, Boston, Portland — have banned government use of facial recognition, but these ordinances don’t cover consumer devices. There is no federal law specifically governing facial recognition technology, despite years of congressional hearings and proposed legislation that has gone nowhere.
Europe presents a different picture. The EU’s AI Act, which began phased implementation in 2024, places significant restrictions on real-time biometric identification in public spaces. Smart glasses equipped with facial recognition capabilities would face substantial regulatory hurdles in European markets. But Meta sells the Ray-Ban Meta glasses in Europe today, and the question of whether a device that enables facial recognition — without directly performing it — falls under the AI Act’s restrictions remains legally untested.
So the burden, for now, falls on Meta’s voluntary choices. And the company’s choices have been telling.
Rather than engaging substantively with the coalition’s requests, Meta has emphasized user-facing controls — the ability to delete your own recordings, to choose whether to share photos, to turn the camera off. These controls address one half of the privacy equation: the wearer’s privacy. They do almost nothing for everyone else’s. The person walking past you on the sidewalk, the barista making your coffee, the child on the playground — none of them have any control over whether the glasses capture their image, where that image goes, or what it’s used for.
Privacy scholars have a term for this: the “third-party problem.” It’s the same issue that plagued Google Glass a decade ago, earning early adopters the derisive nickname “Glassholes.” Google eventually shelved the consumer version of Glass, in part because of the social backlash. Meta is betting that a decade of smartphone ubiquity has normalized constant photography enough that smart glasses won’t provoke the same reaction. So far, that bet appears to be paying off — the Ray-Ban Metas have sold well without generating the visceral public hostility that doomed Google Glass.
But the technology has advanced enormously since 2014. Google Glass had a crude camera and limited connectivity. The Ray-Ban Meta glasses shoot high-resolution photos and video, stream live to Instagram and Facebook, and connect to Meta’s AI systems for real-time visual processing. The potential for misuse isn’t just greater — it’s categorically different.
The coalition’s January letter specifically asked Meta to conduct and publish a human rights impact assessment for the smart glasses line. This is standard practice for technology companies operating under the UN Guiding Principles on Business and Human Rights, which Meta has publicly endorsed. The company has conducted such assessments for other products. Its refusal to do so for smart glasses — or at least its refusal to commit to publishing one — suggests it knows what such an assessment might reveal.
Meta isn’t alone in the smart glasses race. Snap continues to develop its Spectacles line, though it has focused primarily on augmented reality overlays rather than AI-powered camera features. Google is back in the game, reportedly working on new smart glasses that would integrate its Gemini AI. Amazon has its Echo Frames. And a wave of Chinese manufacturers are producing budget smart glasses for global markets. The competitive pressure creates a race-to-the-bottom dynamic on privacy: any company that voluntarily restricts its glasses’ capabilities risks losing market share to competitors that don’t.
This is precisely why the civil society groups are pushing for commitments now, before the market matures and norms calcify. Once hundreds of millions of AI-equipped cameras are sitting on people’s faces, retrofitting privacy protections becomes exponentially harder. The groups understand that Meta, as the market leader with the strongest brand partnership (Ray-Ban’s parent company EssilorLuxottica is the world’s largest eyewear manufacturer), has outsized influence on where the category goes.
Meta, for its part, appears to be calculating that the political environment favors inaction. The current U.S. administration has shown little appetite for technology regulation, and the Federal Trade Commission — which under Lina Khan had been aggressive on privacy enforcement — is operating under new leadership with different priorities. If there was ever a window for the federal government to impose guardrails on facial recognition in consumer devices, it may be closing.
The stakes are not hypothetical. Domestic violence organizations have flagged smart glasses as potential tools for stalkers and abusers — devices that can surreptitiously photograph a partner’s location, contacts, and daily movements without detection. Immigrant rights groups worry about the technology being used for informal immigration enforcement. Civil rights organizations have documented the well-established racial bias in facial recognition systems, which consistently show higher error rates for darker-skinned individuals, raising the specter of misidentification with real-world consequences.
None of these concerns are new. All of them were raised in the coalition’s letter to Meta. And Meta’s response addressed none of them with specificity.
What the company did offer was an invitation to continue the dialogue. More meetings. More conversations. The civil society groups, several of whom spoke to reporters on background, described this as a familiar stalling tactic — the appearance of engagement without the substance of commitment. They’ve seen it before, from Meta and from other tech companies. Talk until the product is too entrenched to change, then point to market adoption as evidence that consumers have accepted the tradeoffs.
The question now is whether external pressure — from regulators, from the press, from consumers themselves — can alter Meta’s calculus before the next generation of smart glasses arrives. The company’s product roadmap reportedly includes models with even more advanced cameras, real-time translation overlays, and deeper AI integration. Each iteration makes the privacy questions more urgent and the answers more consequential.
A tiny LED light won’t cut it. Meta knows this. The civil society groups know this. And increasingly, the public is beginning to figure it out too. The only remaining question is whether anyone with the power to compel change will act before the cameras are everywhere — and the faces of millions are captured by a product designed, above all, to be invisible.
WebProNews is an iEntry Publication