Mark Zuckerberg’s Meta has confirmed that images and videos shared with its AI assistant through the company’s Ray-Ban Meta smart glasses can be used to train the company’s AI models.
TechCrunch reports that Meta has shed light on its data usage practices concerning the Ray-Ban Meta smart glasses. The company has confirmed that any image or video shared with the Meta AI assistant through these glasses may be used to train and improve Meta’s AI models, turning the whole world into Zuckerberg’s dataset.
This revelation comes after TechCrunch inquired about Meta’s stance on using photos and videos captured by users on the Ray-Ban Meta for AI training purposes. Initially, Meta provided a vague response, but later clarified that while images and videos captured on the smart glasses are not used for training by default, once a user submits them to the AI for analysis, they fall under a different set of policies.
The implications of this practice are significant, as Ray-Ban Meta users may unknowingly be providing Meta with a vast collection of personal data, including images of their homes, loved ones, and personal files. This data could potentially be used to develop increasingly powerful AI models. The only way for users to opt out of this data collection is to refrain from using Meta’s multimodal AI features altogether.
Meta’s recent rollout of new AI features for the Ray-Ban Meta glasses has made it easier for users to interact with the AI assistant more naturally, which could lead to an increase in data submissions. The introduction of a live video analysis feature, showcased in a promotional video where a user analyzes their closet to pick an outfit, further emphasizes the potential for data collection.
While Meta’s privacy policy does state that user interactions with AI features can be used to train AI models, the specifics regarding images shared through the Ray-Ban Meta were not initially clear. Meta’s AI terms of service also mention that by sharing images with Meta AI, users agree to have those images analyzed, including facial features, using AI technology.
The use of facial recognition software by Meta has been a contentious issue in the past, with the company recently settling a court case in Texas for $1.4 billion related to its now-defunct “Tag Suggestions” feature on Facebook. It is worth noting that several of Meta AI’s image features are not being released in Texas, possibly due to these legal concerns.
Read more at TechCrunch here.
Lucas Nolan is a reporter for Breitbart News covering issues of free speech and online censorship.
