[ad_1]
Before you open your heart to a cat, your “ride or die big sister” played by Kendall Jenner, or an AI grandpa named Brian on Instagram, know that your messages can’t be private.
Meta’s AI Personas, now in beta, are a collection of characters — some played by celebrities and creators — that users can chat with on Messenger, Instagram and WhatsApp. However, it appears that messages with these characters on Instagram are not end-to-end encrypted.
After chatting with Meta’s AI personalities we have more questions than answers.
With end-to-end encryption off, the option to start AI chat appears.
Credit: Screenshot: Mashable / Meta
After end-to-end encryption is turned off, the option is no longer present.
Credit: Screenshot: Mashable / Meta
In the Messages tab on Instagram, there’s a toggle at the top that allows you to turn on end-to-end encryption, which protects your messages from prying eyes, including meta and government. But when this feature is toggled, the option to start AI chat disappears. If you click the info button (“i” circle icon) in the chat, the “Use end-to-end encryption” option is grayed out. When you click it, a window pops up that says, “Some people still can’t use end-to-end encryption.” It then says that you “can’t add them” – meaning the AI ​​personality – to the chat. You literally have no choice but to communicate with one of these people on Instagram through end-to-end encryption.
This window appears to confirm that Meta’s AI messages are not end-to-end encrypted.
Credit: Screenshot: Mashable / Meta
One of the privacy concerns with the rise of generative AI is the collection of large amounts of data — both to train models and to provide companies with granular insights about their customers. Meta already has a bad reputation for handling personal data. There was the whole Cambridge Analytica scandal, instances of Facebook handing over private conversations to law enforcement, and the way its algorithms leveraged personal data and behavior to make its platforms addictive (and (in some cases harmful) can be made. Past examples suggest that Meta — or any social media company, to be fair — shouldn’t be trusted with your data.
When trying out the AI ​​Messages feature in WhatsApp for the first time, you are immediately presented with a pop-up disclaimer that says, “Meta may use your AI messages to improve AI quality. Yes. But your private messages are never sent to Meta. They can’t be. Read and remain end-to-end encrypted.”
WhatsApp’s disclaimer states that messages are end-to-end encrypted, but this has not yet been confirmed.
Credit: Screenshot: Mashable / Meta
It turns out that, while some information about your messages can be obtained by AI (still not great for privacy), the content of the messages is private. But that’s unconfirmed, especially given Meta’s vague Generative AI privacy policy, which says, “When you chat with an AI, Meta uses the messages you send to train an AI model. could, which would help improve AI.”
Mashable has confirmed to Meta that AI messages on Instagram are not end-to-end encrypted, and also to clarify whether WhatsApp and Messenger are. Although we hadn’t heard back by publication time, we’ll update this story if Meta responds.
Last spring, OpenAI launched an opt-out feature for ChatGPT, which gives users the option to prevent their data from being used for model training. However, other AI chatbots such as Google Bard and Microsoft Bing do not have such opt-out features, although there is the ability to delete your activity. On Meta’s Creative AI Privacy Policy page, there is a similar option to delete your data. You can do this by typing: /reset-ai to remove data from an individual AI chat and by typing: /reset-all-ais to delete data from all chats on Meta apps.
