Meta has quietly launched a new AI feature for Facebook users on iOS and Android, and it’s raising eyebrows. 
This optional feature, called “cloud processing,” allows Meta to access photos and videos from your phone’s camera roll to generate “creative ideas” when posting Stories. If you tap ‘Allow,’ you’re effectively giving Facebook permission to analyze your media with its AI – including facial recognition – and use that data to offer content suggestions.
Although the company claims this feature is opt-in and can be turned off at any time, accepting it also means agreeing to Meta’s AI Terms of Service. These terms allow Meta to store and use your personal media to train its AI, raising privacy concerns once again.
Users who enable this feature may notice Facebook suggesting moments to share, like travel recaps or video collages. To provide these, Meta uploads selected media to its cloud on a recurring basis, analyzing content based on timestamps, locations, detected objects, and people’s faces.
Facebook has confirmed to TechCrunch that this is a limited test currently available in the US and Canada. The company insists the processed content will only be visible to the user unless shared manually. “We’re exploring ways to make sharing easier,” said Maria Cubeta, Facebook’s Video Communications Manager. “Suggestions are opt-in only and not used to train AI models in this test.”
However, many remain skeptical given Meta’s history of data collection. While the company continues its push toward “superintelligence” through its Meta Superintelligence Labs, this move may feel more like another step into your private gallery than a leap toward innovation.