Meta AI App Users Unwittingly Expose Chatbot Logs Publicly

Users of Meta Platforms Inc.’s new artificial intelligence app have accidentally made their chatbot conversations public, exposing sensitive information like addresses and court case details. This issue was highlighted by TechCrunch and traced back to the Meta AI app, launched in late April as a standalone version of a chatbot previously integrated into Meta’s social media platforms. The app allows users to generate responses, search the web, and create images, with personalization based on their account data. A “Share” button enables users to post chat responses to a public Discovery feed, but many remain unaware of this feature’s implications. Although the app’s download count is relatively modest at 6.5 million, the potential privacy breach raises concerns, especially given Meta’s history with regulatory fines over privacy issues due to similar design choices. This incident could draw further scrutiny towards Meta’s AI strategies.

Source link