Ray-Ban Meta smart glasses put a camera on the wearer’s face and connect that camera to AI features. The central privacy question is simple: after those images are captured and uploaded for AI processing, can they become training data?
When TechCrunch asked Meta that question, the company did not give a yes or no answer. That silence matters because the glasses can take photos intentionally, but they can also capture images when AI features are activated by certain keywords such as "look".
What Meta Would Not Confirm
TechCrunch asked Meta whether it plans to train AI models on images from Ray-Ban Meta users, as it does with images from public social media accounts. Meta did not commit to keeping those images out of AI training.
In a video interview with TechCrunch on Monday, Anuj Kumar, a senior director working on AI wearables at Meta, said, "We’re not publicly discussing that," when asked about the issue.
Meta spokesperson Mimi Huggins, who was also on the call, said, "That’s not something we typically share externally." When TechCrunch asked directly whether Meta is training on these images, Huggins said, "we’re not saying either way."
The distinction is important. A photo taken through smart glasses is not the same as a public social media post. It can show a room, a closet, nearby people, personal belongings, or whatever happens to be in front of the wearer when an AI feature is running.
Why Passive Capture Changes The Stakes
The privacy issue is not only about photos a user deliberately chooses to take. Ray-Ban Meta’s AI features can also trigger camera use through certain keywords. That means the glasses may capture images during an interaction where the user is focused on asking a question, not on building a photo record.
TechCrunch previously reported that Meta plans to launch a new real-time video feature for Ray-Ban Meta. When activated by certain keywords, the glasses will stream a series of images into a multimodal AI model. The goal is to let the system answer questions about the user’s surroundings in a low-latency, natural way.
In practical terms, that kind of feature can involve many images. The source describes a scenario in which a user asks the glasses to scan the contents of a closet to help choose an outfit. The glasses would effectively capture dozens of photos of the room and its contents, then upload them to an AI model in the cloud.
That workflow raises a direct follow-up question: what happens to those photos after the AI model uses them? Based on the source, Meta would not say.
The Public Data Line Gets Harder To Draw
Meta has already said it is training its AI models on every American’s public Instagram and Facebook posts. The company treats that material as "publicly available data", according to the source.
Smart glasses make that logic more difficult. A user’s field of view can include private spaces and people who did not choose to be involved in the interaction. The source argues that the world seen through smart glasses is not obviously "publicly available" in the way a public post is.
This is where Ray-Ban Meta becomes more than a product question. It becomes a boundary question for AI wearables:
- Is an image private because it was captured in a personal setting?
- Does AI processing change how long that image may be kept or reused?
- Should images from smart glasses be treated differently from public social media posts?
- What should bystanders be able to expect when someone near them is wearing a camera?
The source does not provide Meta’s answers to those questions. The key fact is that Meta declined to publicly say whether the images are used for AI training.
Why Face Cameras Remain Sensitive
Ray-Ban Meta smart glasses also revive an older social concern: many people are uncomfortable around face-worn cameras. The source points to Google Glass as an example of how visible or wearable cameras can create unease among people nearby.
That discomfort is easy to understand from the facts described. The camera is on the wearer’s face, and the AI system can capture images of the surrounding world. Some of those images may be taken intentionally, while others may be collected as part of an AI interaction.
A clear privacy promise could reduce uncertainty. The source suggests that Meta could say the photos and videos from the glasses will remain private and isolated to the device’s camera experience. But according to the article, Meta is not making that commitment here.
Other AI model providers are described as having clearer rules for user data. Anthropic says it never trains on a customer’s inputs into, or outputs from, one of its AI models. OpenAI also says it never trains on user inputs or outputs through its API.
What Users Can Take Away
The immediate takeaway is not that Meta confirmed it trains on Ray-Ban Meta photos. It did not. The takeaway is that Meta also would not rule it out in the exchange described by TechCrunch.
That difference matters for anyone considering AI-powered smart glasses. A device that can see what the user sees may collect sensitive visual context, especially when AI features passively capture images during a request.
Until Meta gives a clearer public answer, the privacy status of those images remains uncertain based on the source. For a product category built around cameras, AI models, and the wearer’s immediate surroundings, uncertainty is itself a major part of the story.