The recent implementation of artificial intelligence (AI) characters in Goal has generated a debate around the security and privacy of messages. Even though these AI characters on Messenger, Instagram and WhatsApp promise a new way of interaction, a report suggests that messages on Instagram lack end-to-end encryption.
As it explains Masheable, in the Instagram messages tab in its beta version, users can activate the end-to-end encryption feature to ensure the security of their conversations. However, when this feature is activated, the option to start a chat with an AI character disappears.
If users try to start a conversation with an AI character in an encrypted chat, they are informed that “some people cannot yet use end-to-end encryption” and therefore cannot add the AI person to the chat. chat.
Data privacy concerns
The lack of encryption in these chats with AI characters raises additional data privacy concerns. Meta has faced privacy issues in the past, including the Cambridge Analytica case, handing over private conversations to authorities and using personal data to increase addiction to its platforms. These past incidents suggest that trust in Meta when it comes to data privacy is questionable.
The disclaimer that appears when using the AI messages feature in WhatsApp suggests that Meta may use these messages to improve the quality of the AI, but ensures that personal messages are not sent to Meta and remain end-to-end encrypted.
However, Meta’s privacy policy regarding generative AI is ambiguous and suggests that messages sent to the AI may be used to train the AI model.