24/09/2023
iOS 17 is out. Amongst other things, it now contains ‘Personal Voice’, a feature that uses AI to clone your voice. You can then type phrases that it will read out in your voice, either through the phone’s speakers or in FaceTime and other apps.
I’ll be honest, I’ve got concerns about this kind of technology for all sorts of reasons. Apart from the fact it’s a bit creepy, I’d have thought its potential misuse for fraud etc would be a big issue.
Then there’s the security issue. I’ve been doing some research into Apple’s various privacy docs recently and, to be honest, the company is pretty good at securing your data, giving you access to it and providing often detailed information about how stuff works. In the case of Personal Voice, all processing is completed on your device rather than in the cloud but, like other data, it can be shared with other devices over iCloud. Apple uses end to end encryption to store it so it is only accessible on your PIN protected devices and even Apple can’t access it. Even so, if, like me, you’re not keen on storing such personal information on a company’s servers, you can disable ‘share across devices’ ideally before setting it up.
The real question, though, is ‘What’s the point?’.
If you have a disability that makes speaking difficult then this could be useful. If not, I suspect this feature could be a source of more problems than solutions.
For more info, Mozilla (always an excellent source of reliable information) have this to say…
Apple will let you make a copy of your voice in iOS 17, its latest software for the iPhone. Here’s how you can use Apple’s AI to make a copy of your voice. But should you?