How to Stop Meta Training AI Models on Your Smart Glasses' Visual Data

Users can adjust this preference in the device’s companion app under privacy

Meta has introduced a privacy setting that controls whether the company can use images and video captured by its Ray-Ban Stories smart glasses to train artificial intelligence models. This update comes amid growing scrutiny over how tech companies collect and utilize personal data from wearable devices. The setting directly determines whether Meta’s contractors or automated systems can access photos and videos recorded through the glasses for purposes such as improving object recognition or scene understanding.

Latest news

Why Meta's New AI Tool Fails to Win Over Daily UsersAssessing Risks in Open-Source Artificial IntelligenceAutonomous AI Agents Launch Cyberattacks Against North American Government PortalsWarlock Group Targets SharePoint Servers in Critical Infrastructure

Users can adjust this preference in the device’s companion app under privacy controls, a change Meta says was made in response to user feedback and regulatory pressure. How the Data Review Process Actually Works When users opt in to data sharing, Meta may send anonymized clips to third-party vendors where workers label elements like faces, landmarks, or objects to help train machine learning algorithms. These reviewers do not see account information but can view the raw visual content. Meta claims the process is designed to improve features like live translation and augmented reality overlays, though critics argue the lack of transparency makes informed consent difficult.

The company says it retains such data for no longer than six months unless

The company says it retains such data for no longer than six months unless explicitly permitted otherwise. Can You Really Trust Meta to Honor Your Choice? Despite the new toggle, privacy advocates question whether opting out fully prevents data use, pointing to past instances where similar settings had loopholes or were enabled by default. Meta insists the setting is respected across its systems and that disabling it stops both human review and AI training use of the visual feed. However, the company continues to collect other types of data from the glasses, including voice commands and interaction metrics, which may still contribute to model improvement indirectly. Users are advised to regularly check their settings, as software updates could reset preferences. Frequently Asked Questions Does turning off this setting delete previously collected data? No, disabling the setting only stops future use of your visual data for AI training or human review.

Data already collected prior to the change may still be retained unless separately deleted through your account settings. Will opting out affect the functionality of my smart glasses? Meta states that core features like photo capture, video recording, and basic AR functions will remain unaffected. However, some personalized experiences that rely on AI trained on visual patterns might be less accurate over time if insufficient data is contributed. Are workers still able to see my data if I opt out? According to Meta, no. When the setting is disabled, your photos and videos are not sent to human reviewers or used in any training pipelines, whether automated or manual.