Meta announced plans to add its Private Processing technology to Ray-Ban Meta smart glasses at its annual Connect event in Menlo Park, California, according to a report from Wired. The feature, which the company first built for WhatsApp last year, aims to let the glasses' AI assistant personalize responses and handle requests without Meta gaining access to any user information. The technology represents Meta's effort to address mounting privacy concerns around its wearable devices.
Private Processing works by enabling AI cloud processing that shields data from outside parties, including Meta itself, the report notes. For smart glasses, the system could theoretically support private chat interactions with Meta's AI assistant, as well as transcriptions and messages, though Meta hasn't yet specified which features it will actually support. The company introduced Private Processing into WhatsApp and its Meta AI app last year, where it helped power WhatsApp's Incognito Chat feature. The aim was to allow customers to use AI services without compromising WhatsApp's end-to-end encrypted secure messaging.
Meta positioned the announcement as sharing "excitement about this enhanced privacy platform," with Pritam Shah, director of engineering, telling Wired that "we have the platform ready. We have it working." Shah emphasized the company is announcing the technology now, before launching specific experiences, so it can collaborate with the security community, privacy community, and third-party auditors. The company described AI glasses as having "the capacity to provide superpowers like translating a conversation or summarizing notes or a conversation," according to Meta's press release.
The timing reflects a challenging moment for Meta's relationship with users over privacy issues. This year alone, the company has weathered months of backlash over its face-recognition technology and concerns that its smart glasses could enable sexual predators, with some dubbing them "pervert glasses" and building apps like ZuckOff to detect the devices nearby. Years of security and privacy missteps have damaged user trust, and Meta's products have often appeared more effective at monitoring users than assisting them. The company is now working on Confidential VM, an enhanced version of its Secure VM cloud environment for the Muse AI agent, which promises similar data inaccessibility guarantees and will also come to Meta's glasses.
Private Processing isn't available yet on Meta's glasses, and Shah doesn't have a timeline for when it will arrive or which specific features will work with it. When asked about examples the company shared, Shah clarified "these are not concrete features coming, but these are some examples. And there will be even more powerful examples." The company appears to be betting that transparency about privacy safeguards before launch will help rebuild credibility with users increasingly wary of data collection on wearable devices. Whether announcing privacy features before they're functional will satisfy critics or simply highlight the gap between Meta's promises and its current capabilities remains an open question for a company whose track record has left many users skeptical of assurances alone.