Meta Adds Visual Data AI Opt-Out for Smart Glasses but Keeps Audio Training
The new setting blocks camera-captured visual data from AI training, while audio interactions can still be used and manually deleted.
- On Monday, Meta announced users can opt out of having 'visual data' from smart glasses used for AI training, preventing camera-based interactions from being reviewed by third-party contractors.
- Earlier this year, reports revealed third-party contractors reviewed intimate user images from the glasses, while dormant face-recognition code called NameTag reached an app downloaded more than 50 million times, sparking privacy concerns.
- At Meta Connect on September 23, 2026, Meta introduced Private Processing, an 'enhanced privacy platform' using specialized encrypted servers where nobody, including Meta, can access user data.
- Pritam Shah, Meta's director of engineering, confirmed the platform is ready and running safety tests, with the Bug Bounty program expanded to audit the technology, though it hasn't launched for glasses yet.
- By the end of 2026, Meta plans to exceed 100 glasses options, including camera-free Ray-Ban Meta Audio, soon integrating the Muse personal AI agent for expanded AI-rich experiences.
13 Articles
13 Articles
Meta smart glasses have a new AI-training opt-out
Meta has changed the privacy settings on its glasses. Now, users can opt out of sending visual data to train AI.
Called Private Processing, resource uses encrypted servers to process information captured by glasses without allowing Meta itself access to data
New Meta Glasses Pitch an 'Enhanced Privacy Platform' but Face-Recognition Code Already Hit Millions of Phones
Meta is pitching its expanding smart-glasses line on privacy just months after face-recognition code for its glasses reached an app downloaded more than 50 million times. At Meta Connect on 23 September 2026, the company said it plans to bring its Private Processing technology to its smart glasses, describing it as an 'enhanced privacy platform' designed to let AI services process sensitive information without Meta itself accessing the data. Met…
Meta Is Bringing Encryption to Its Smart Glasses to Keep AI Chats Private
Meta sees a future where users will converse daily with AI using its smart glasses. So, it's building Private Processing for the specs to ensure that only the wearer has access to the data. To keep AI conversations private, Meta is laying the groundwork to expand its encryption technology to smart glasses, in anticipation that most consumers will eventually interact with AI through the wearables.The company mentioned the effort at Meta Connect, …
Meta Adds Visual Data AI Opt-Out for Smart Glasses but Keeps Audio Training
Meta released its new personal AI agent, Muse, in September 2026. The launch of “Muse” attempts to address privacy concerns regarding how wearable cameras interact with the physical world. For users who opt out, visual data used for real-time AI tasks, such as identifying a landmark or translating text on a sign, is processed once and then immediately deleted rather than stored. Under the new opt-out setting, visual data is processed once and im…
Coverage Details
Bias Distribution
- 83% of the sources lean Left
Factuality
To view factuality data please Upgrade to Premium











