Meta AI App Introduces Warning Against Sharing Personal Information

Meta has taken a significant step to address the growing concern of over-sharing in the public feed of its AI application. The company has implemented a new disclaimer that advises users to “avoid sharing personal or sensitive information” when using the “post to feed” feature in the Meta AI app.

This change was first reported by Business Insider, which described the app as “one of the most depressing places online” due to the overwhelming amount of intimate, embarrassing, and sometimes personally identifiable information that users were sharing publicly, often without realizing the implications. Although Meta AI does not share users’ chat histories by default, many users appeared to be unaware that their interactions could be made visible to the public through the app’s built-in “discover” feed.

Recent observations revealed posts where users sought advice on topics such as “improving bowel movements” and whether a relative could be held liable for their employer’s unpaid taxes. In one instance, a user attempted to mitigate embarrassment by stating “keep this private” in a public post, highlighting the confusion surrounding the app’s sharing capabilities. Such unusual public interactions have been occurring since the Meta AI app’s launch in April, but they gained renewed attention last week as social media users began discussing the peculiar conversations visible in the app’s “discover” feed.

Privacy experts have criticized Meta for this feature, noting that most mainstream AI chatbots do not include a social, publicly visible feed. Rachel Tobac, a security expert who has previously collaborated with Meta, stated, “If a user’s expectations about how a tool functions don’t match reality, you’ve got yourself a huge user experience and security problem.” She emphasized that users have developed a schema around AI chatbots and do not expect their prompts to appear in a social media-style discover feed, which is not the norm for other tools.

The Mozilla Foundation also urged Meta to reconsider the app’s design, pointing out that “Meta AI’s app doesn’t make it obvious that what you share goes fully public.” They highlighted the lack of clear iconography and familiar cues about sharing, which are present in other Meta applications.

In response to the feedback, Meta has introduced a warning that appears when users choose to share a Meta AI interaction publicly. This warning, however, only seems to appear on the first instance of sharing. It states, “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.” This change aims to enhance user awareness regarding the visibility of their shared content.

As reported by Business Insider, the app’s public feed now appears to exclude text exchanges shared by other users, focusing instead on AI-generated images and videos. It remains unclear whether this alteration is a permanent change or a temporary response to the recent scrutiny the app has faced. Meta has been contacted for further clarification, and updates will be provided if a response is received.

For users who have inadvertently shared personal posts publicly, there is a way to remove them. By tapping on their profile in the top right corner of the app, users can navigate to Data & Privacy -> Manage your information -> Make all public prompts visible only to you and select “apply to all” to ensure their prompts are kept private.

This article originally appeared on Engadget at spotted.