According to Techcrunch, Facebook has recently been requesting permission from users to access the user’s mobile phone album to provide artificially intelligent editing advice. Notablely, this feature targets not only images already uploaded by users, but even private images that have never been shared with the platform.
This feature usually pops up when the user creates new content. The screen will request the user to choose whether to add the ” cloud-processing ” option to get ” creative advice ” . Once a user is allowed to click, it means that Facebook is authorized to generate new ideas from its mobile album. This may include photo collages, annual reviews, AI stylish processing or thematic photo collections.
Facebook notes that for this feature to be effective, documents in user albums will be uploaded to the server on an ongoing basis, based on information such as time, place or theme. Facebook further stated that by agreeing to the Meta AI service clause, users agreed with AI to analyse their media files (including facial features). The company will use information on the date, person or object in the image to generate creative ideas.
Meta’s AI clause clearly states: “When photographs are shared (upload to cloud) you agree that Meta will use AI to analyse these images, including facial features. This processing enables us to provide innovative new features, including the ability to summarize the image content, modify the image and generate new content based on the image.”
The same article also provides that Meta AI has the right to “reserve and use” any personal information shared by users in order to individualize its AI output. The article also states that Meta may review the user ‘ s interaction with AI, including the content of the dialogue, and that the review will be completed manually. It was reported that Meta did not provide a clear definition of what was meant by “personal information”, but merely stated in general terms to include “information submitted by you in the form of tips, feedback or other content”.
Notably, Meta had previously confirmed that it had trained its generation AI model using publicly available data published on Facebook and Instagram since 2007. The new functionality, which involved access to undisclosed photographs on user’s private equipment and cloud processing, raised broader privacy concerns. The functionality connects the user ‘ s local photo bank with a strong cloud-based AI analytical capacity, and its data collection and use boundaries draw attention to users and privacy advocates.
