Meta has rolled out a privacy warning after users unintentionally exposed sensitive conversations on its AI app, following backlash from privacy advocates and the public. 
The new disclaimer alerts users that anything they post could become visible to everyone, urging them not to share personal or confidential content.
This move comes amid a broader industry push for ethical AI practices and tighter data privacy standards. As AI platforms like Meta’s become more widespread, many users assume their interactions are private, leading to unintentional oversharing of intimate details. From tax issues to health questions, such private content has been turning up on Meta AI’s public ‘Discover’ feed – often without users realizing it.
The issue, highlighted by Business Insider, has led to criticism that Meta’s app design encourages public visibility by default, unlike other AI platforms. Some conversations have been described as deeply personal, and the lack of clear notification sparked concern across social media when screenshots of odd prompts began circulating online.
Security expert Rachel Tobac, who has worked with Meta in the past, emphasized the disconnect between user expectations and reality. “People don’t anticipate their chatbot prompts showing up like social media posts. That’s just not how AI tools usually work,” she noted.
Privacy watchdog Mozilla Foundation also called on Meta to rethink its interface, demanding more transparent user alerts when content is made public. In response, Meta issued a one-time warning label informing users: “Prompts you post are public and visible to everyone. Your prompts may be suggested by Meta on other Meta apps. Avoid sharing personal or sensitive information.”
While this disclaimer is a step forward, experts argue it doesn’t fully address the underlying design flaw. Many believe Meta should revamp the app’s structure altogether to prioritize user privacy by default. Until then, users are urged to think twice before treating AI chats like a private space.