Ray-Ban Meta smart glasses are becoming a clearer test case for how wearable cameras and consumer AI can reshape privacy expectations. Meta has confirmed that when users share images or videos with Meta AI, that material may be used to improve its AI systems in places where multimodal AI is available.
The distinction matters. A photo taken on the glasses is treated differently from a photo sent to Meta AI for analysis. Once a user asks the assistant to interpret what the camera sees, the image moves into a separate policy area with broader implications for AI training.
What Meta Says Happens To Shared Images
Meta told TechCrunch that images and videos shared with Meta AI may be used to improve its AI in locations where multimodal AI is available, currently US and Canada. Meta policy communications manager Emil Vazquez said in an email to TechCrunch: "[I]n locations where multimodal AI is available (currently US and Canada), images and videos shared with Meta AI may be used to improve it per our Privacy Policy,"
A Meta spokesperson also clarified a narrower point: photos and videos captured on Ray-Ban Meta are not used by Meta for training as long as the user does not submit them to AI. That means capture alone is not the trigger described in the source. The trigger is sharing the material with Meta AI.
That creates a practical rule for users: if they want to avoid sending an image into Meta AI’s training pipeline, the only described option is not to use the multimodal AI features. The source describes this as the only way to “opt out” of that specific use.
Why The Smart Glasses Context Raises The Stakes
The concern is not only that Meta AI can analyze an image. It is that Ray-Ban Meta glasses are worn on a person’s face and can capture what the wearer is looking at. The images a user asks AI to analyze may include homes, loved ones, personal files, or other private surroundings.
Meta spokespeople told TechCrunch that the Ray-Ban Meta user interface makes this clear. But the source also notes that company executives either initially did not know or did not want to share these details with TechCrunch. That gap matters because privacy expectations depend on users understanding what happens before they use the feature.
The broader issue is how Meta defines useful AI training data. The source notes that Meta already trains its Llama AI models on everything Americans post publicly on Instagram and Facebook. With Ray-Ban Meta, the concern is that data shared through smart glasses can extend the category of AI-relevant material to things people look at and ask a chatbot to analyze.
New AI Features Could Increase Data Sharing
The timing is important because Meta started rolling out new AI features on Wednesday that make it easier for Ray-Ban Meta users to invoke Meta AI in a more natural way. If the assistant becomes easier to call on during everyday use, users may send more images and videos to Meta AI.
Meta also announced a new live video analysis feature for Ray-Ban Meta during its 2024 Connect conference last week. According to the source, the feature essentially sends a continuous stream of images into Meta’s multimodal AI models.
In a promotional video, Meta presented a use case where someone could look around a closet, have AI analyze the whole thing, and pick an outfit. The privacy issue is what sits behind that convenience: the user is also sending those images to Meta for model training, according to the source’s interpretation of Meta’s statements and policies.
The Policies Behind The Feature
Meta spokespeople pointed TechCrunch to the company’s privacy policy, which states: "your interactions with AI features can be used to train AI models." The source says this appears to include images shared with Meta AI through Ray-Ban smart glasses, although Meta still would not clarify that point further.
Spokespeople also directed TechCrunch to Meta AI’s terms of service. Those terms state that by sharing images with Meta AI, "you agree that Meta will analyze those images, including facial features, using AI." That language is especially significant for a device that can capture people and places in ordinary settings.
The source also points to Meta’s recent legal history in Texas. Meta paid the state of Texas $1.4 billion to settle a court case connected to its use of facial recognition software. The case concerned a Facebook feature rolled out in 2011 called "Tag Suggestions." By 2021, Facebook made the feature explicitly opt-in and deleted billions of people’s biometric information it had collected.
Several of Meta AI’s image features are not being released in Texas, according to the source.
Voice Data Adds Another Layer
The Ray-Ban Meta privacy picture is not limited to images. The source says Meta stores all transcriptions of voice conversations with Ray-Ban Meta by default to train future AI models.
For actual voice recordings, the source says there is a way to opt out. When users first log in to the Ray-Ban Meta app, they can choose whether voice recordings may be used to train Meta’s AI models.
Taken together, the image, video, and voice details show how smart glasses can become a broad input channel for AI systems. The device is not just a camera or a headset. It is a wearable interface that can feed visual and voice interactions into AI features.
The Larger Smart Glasses Privacy Question
Meta, Snap, and other tech companies are pushing smart glasses as a new computing form factor. These devices put cameras on people’s faces and increasingly rely on AI to make sense of what those cameras see.
That brings back privacy concerns associated with the Google Glass era. The source also notes that 404 Media reported some college students had already hacked Ray-Ban Meta glasses to reveal the name, address, and phone number of anyone they look at.
The central issue is straightforward: wearable AI can make analysis feel effortless, but that ease may obscure what is being shared. For Ray-Ban Meta users in US and Canada, asking Meta AI to analyze an image or video can mean contributing that content to systems designed to improve Meta’s AI.