Meta AI chatbot users are largely sharing intimate conversations with artificial intelligence applications without realizing it.
The Washington Post turned its attention to this issue and recalls a few months ago that Meta launched another app for an artificial intelligence-based chatbot to provide users with personalized answers to the question.
However, many users do not take into account specific nuances. Unlike ChatGPT and many other artificial intelligence applications, Meta AI has unique features. «Public feed» that users can publish to publish personal conversations with Meta AI.
Meta spokesman Daniel Roberts explained that chats with Meta AI are set to private by default, and users must click the “Share” or “Publish” button before they can be seen in the app's public feed.
At the same time, the “Share” button doesn't directly tell users where the conversation with meta AI will be exposed or if others can see it, so many users may not understand the outcome.
Therefore, the app's public feed was characterized by many intimate conversations about relationships, sex life, faith in God, or financial difficulties.
«We've seen many examples of sending very personal information to AI chatbots and saying something very intimate,” said Callie Schroeder, senior advisor at the Center for Electronic Privacy Information.
She expressed the view that many people assume there is a basic level of privacy, but in reality it is not.
The publication also reports that questions asked to meta AI in the public domain are of inappropriate nature, such as creating pornographic content using AI.
We've written about interesting stories before, as the meta AI chatbot was confused about the US president. After Donald Trump took office, the Meta AI Chatbot still claimed that Joe Biden was president.
Amid concerns about Chinese startup DeepSeek, Meta is known to be planning to invest $60 billion in AI development in 2025.
