ChatGPT Now Controls iMessage, Raising Apple Privacy Concerns

Digital security researchers have revealed that the ChatGPT model is now capable of controlling the iMessage application on Apple devices via the "Computer Use" feature, which allows artificial intelligence to interact with graphical user interfaces just like a human.

Privacy experts believe this capability opens the door to alarming scenarios, as it enables AI models to read, reply to, and delete private messages without the user always realizing the extent of access granted. Apple has not yet issued an official comment on these developments.

These developments come amidst a race among AI companies to expand the capabilities of their models to control applications and devices, while governance and regulatory frameworks lag behind this rapid growth. Reports indicate that this feature is currently available in limited beta versions, but its widespread adoption is now merely a matter of time.

This evolution embodies a difficult trade-off facing users: leveraging AI capabilities to manage their digital lives in exchange for giving up layers of privacy that were protected until yesterday. The irony is that Apple built its brand identity on privacy, while the doors of its devices are now being opened to external models operating under a completely different logic.

What Do These Terms Mean?

Computer Use: A capability that grants AI models the authority to control app interfaces, buttons, and text on a computer screen — as if a real user were sitting in front of the device.

iMessage: Apple's instant messaging app, featuring end-to-end encryption between the two parties of a conversation, making any external access to it security-sensitive.

Large Language Models (LLMs): Artificial intelligence systems trained on vast amounts of text to understand and interact with language — with ChatGPT being one of the most prominent.