Security experts have raised a new concern regarding Apple Intelligence. While Apple takes great measures to keep your data secure, there might be a loophole when it relies on another AI service, ChatGPT. We just fixed one a couple of days ago where the ChatGPT macOS app exposed user conversations in plain text.
But first, here’s how Apple Intelligence prioritizes privacy:
- Thinking on your device first: Whenever possible, Apple Intelligence processes your requests directly on your device, keeping your data entirely private.
- Apple’s secure cloud when needed: If your request requires more processing power, Apple uses its own secure cloud servers, keeping your data within Apple’s control.
- ChatGPT as a last resort: Only when both on-device processing and Apple’s servers can’t handle the request will Apple ask your permission to use OpenAI’s ChatGPT service.
Sounds secure, right? Well, there’s a catch. Security experts like Bruce Schneier from Inrupt, which is a privacy-focused company, believe there might be a privacy risk when using ChatGPT.
Apple has put together a “pretty impressive privacy system” for its AI, says Schneier. “Their goal is for AI use—even in their cloud—to be no less secure than the phone’s security. There are a lot of moving parts to it, but I think they’ve done pretty well” […]
For some data, “you’re stuck with OpenAI’s rules,” says Schneier. “Apple strips identifying information when sending OpenAI the queries, but there’s a lot of identifying information in many queries.
As reveled by 9to5Mac
Imagine asking for a personalized travel itinerary. The details you provide, like the specific city and interests, could reveal your identity, especially to someone who knows you well.
So, while Apple Intelligence offers strong privacy features, it’s recommended that you be mindful when using ChatGPT for particular personal requests.