Smart Glasses Sales Nearly Doubled. The Privacy Backlash Just Doubled Down.

Smart glasses just had their strongest sales quarter on record while the privacy fights around them are leaving the op-ed page and entering federal letters, courtrooms, and criminal dockets. EssilorLuxottica, the Italian-French eyewear conglomerate that makes Ray-Ban and Oakley frames, told investors last week that its smart-glasses segment revenue nearly doubled year over year in the second quarter of 2026 even as overall company revenue declined (GlobeNewswire via Manila Times, Vision Monday). A category that was a side bet for a luxury eyewear house a year ago is now the only part of the P&L accelerating, and the friction around it is accelerating with it.

Meta, EssilorLuxottica's partner on the Ray-Ban Meta line, sold roughly 7 million pairs of its AI-equipped glasses in 2025, according to figures the company cited in its February earnings cycle (CNBC). The category has crossed the threshold where the documented harms start to compound.

The harms are documented, not theoretical. A Svenska Dagbladet (Swedish daily newspaper) investigation found that Meta contractors were shown sensitive footage captured by the glasses, including people having sex, using the bathroom, and entering banking information, according to a Gizmodo summary of the reporting. The investigation also reported a separate case in which a contractor allegedly used a smart-glasses recording to attempt extortion. The data pipeline behind those captures is not hidden. Meta's terms of service reserve the right to use video and photos triggered by the "Hey Meta" command to train its AI models, and that clause is unchanged. Anyone who uses the wake phrase is, by default, opting in to having that clip become training data.

The same hardware is now showing up in places that have nothing to do with the marketing pitch. Smart glasses have also appeared as evidence and as recording devices in U.S. court proceedings, a category of misuse that judges are still writing rules for. Each incident is a single data point. At seven million units a year, the data points start to add up.

The next rulebook probably will not be written by Meta. Federal lawmakers and privacy groups have already pushed back on the company's reported plan to add facial-recognition features to the glasses, a feature that would let the hardware do the work that today still requires a human reviewer, according to industry reporting. The contractor-exposure reporting and the courtroom cases all give rule-makers a paper trail.

The volume of units in the wild is about to get larger. Google and Samsung are both preparing consumer smart-glasses products for release in 2026, according to industry reporting. Two more mainstream brands shipping means more devices on more faces, more captures, more contractor exposure, more courtroom disputes, and more pressure on the "Hey Meta" data-training clause. The original Glasshole moment in 2013 happened when one company shipped a few thousand units to a niche developer base. This round starts with seven-figure annual sales and two new competitors already on the launchpad.

The next twelve to twenty-four months will be settled in public, on three specific fronts: the data-training terms inside the existing glasses, the facial-recognition feature Meta wants to ship, and the social and courtroom norms that have to be invented in real time. At a London show earlier this summer, the singer Lorde paused between songs and told the crowd to take their smart glasses off if they were wearing any. It was a small moment in a long fight. The friction she was pointing at is the part that does not need a regulator to settle. The privacy suits and the lawmaker letters will take longer. The adoption is happening now.