Recently, Facebook, owned by social media giant meta, is testing a new feature aimed at optimizing images uploaded by users through artificial intelligence (AI) technology, thereby enhancing users’ interactive experience on the platform. According to TechCrunch and other technology media reports, this test mainly revolves around when users create Story, triggering a new round of discussions on user data privacy and AI ethics.
Specifically, when a user tries to create a new Story on Facebook, a dialog box will pop up asking the user if he is willing to join the so-called “creative suggestions” program and prompt the user to authorize Facebook to access its mobile album. This prompt design is simple and intuitive. Users can turn on this function by simply clicking “Allow”. This design cleverly reduces the time cost and cognitive burden of users and increases the possibility of authorization.
After the user’s authorization, Facebook will use AI technology to analyze the user’s album content and generate various creative content, including collages, photo review videos, AI-style converted pictures, and photo collections of specific topics. meta explained that to achieve this goal, user media files are regularly uploaded to their cloud servers and organized and processed based on tags such as time, place or topic. This cloud-based AI processing method can effectively utilize meta’s powerful computing resources to provide faster and more diverse creative generation.
While these creative suggestions seem attractive, while authorizing access to the album, it actually means agreeing to meta’s AI Terms of Use. These terms give meta’s AI system permission to analyze user media files, including but not limited to facial feature recognition, and creative concepts based on elements such as dates, characters or objects in the photo. meta stressed that these suggestions are for personal viewing by users and will not be used for advertising targeting. However, this statement does not completely eliminate user concerns.
In this regard, TechCrunch pointed out that this move by meta may objectively prompt users to share more album images, thereby providing meta with more data to train its AI models. The AI Terms of Use clearly state that meta’s AI has the right to “retain and use” any personal information shared by users in order to personalize the output of AI. The ambiguity of this clause has raised concerns about the privacy and security of user data. Although meta promises not to use this data for advertising, the risk of data breaches and abuse still exists.
More noteworthy is that the terms state that meta has the right to review user interactions with AI, including conversation content, and that these reviews may be done manually. However, the terms do not clearly define the specific definition of “personal information” and only generally state that it covers information submitted by users as prompts, feedback or other content. This ambiguity increases users’ uncertainty about how their own data is used and protected. While social media platforms use AI technology to provide services, they must pay more attention to the protection of user data privacy and formulate more transparent and complete terms of use to win the trust of users.
