Meta warns consumers to think twice before buying or using its Ray‑Ban smart glasses

Meta’s Ray‑Ban Display Glasses have finally crossed the threshold from boutique gadget to a product that many consumers will encounter on the street, but the shift brings a host of privacy and civil‑rights questions that have long lingered in the smart‑glasses arena. According to Eff, the glasses now store captured media locally only until the user imports it to a phone, at which point the Meta AI mobile app automatically syncs the files to the cloud. The app is required for setup and for most users to download footage, meaning that, in practice, almost every recording is routed through Meta’s servers. Eff notes that none of the AI features run on‑device; any command such as “Hey Meta, start recording” sends the audio and video stream to Meta for processing, and a portion of that data is used to train the company’s generative models.

The data pipeline does not stop at Meta’s own infrastructure. Eff reports that Swedish newspaper investigations have uncovered human reviewers annotating footage that includes highly sensitive content—nudity, sexual activity, and bathroom use. Meta has told the BBC that such reviews are “in accordance with its terms of use,” which explicitly allow the company to inspect interactions with its AI, whether automated or manual. Moreover, once the media is saved to the phone’s camera roll via the Meta AI app, it may be uploaded to Apple or Google clouds depending on the user’s sync settings, exposing the files to additional corporate custodians and, potentially, law‑enforcement requests.

The privacy concerns are amplified by the glasses’ form factor. Eff observes that the devices look like ordinary eyewear, making it difficult for bystanders to discern when they are being recorded. This “stealth” capability differentiates smart glasses from smartphones, which are conspicuous when filming, and raises the risk of covert surveillance in public and private settings. The same report warns that the default recording behavior—press the button, capture, then automatically upload—leaves users with limited control over what is retained, especially since audio from voice‑activated AI interactions is saved by default and must be manually deleted after each use.

Meta’s dominance in the market is already being challenged. While the company’s partnership with Ray‑Ban and Oakley has made its glasses the most visible, Eff points out that Google has announced a collaboration with Warby Parker on “AI‑powered smart glasses,” and rumors swirl about an Apple entrant. The competitive pressure could accelerate feature rollouts, but it also underscores the urgency of establishing clear data‑handling standards before the technology becomes ubiquitous.

Finally, the broader regulatory context cannot be ignored. The European Union’s forthcoming AI Act and various U.S. state privacy statutes are likely to scrutinize devices that continuously capture audio‑visual data and feed it to large‑scale AI models. Meta’s existing practices—automatic cloud sync, human review of raw footage, and cross‑company data sharing—could place the company at odds with emerging legal frameworks, prompting both lawmakers and civil‑rights groups to demand greater transparency and opt‑out mechanisms. As the glasses move from curiosity to commonplace, consumers will need to weigh the convenience of hands‑free recording against the reality that their most intimate moments may be stored, analyzed, and possibly disclosed far beyond the moment of capture.