Let's imagine that I leave a Mac with wireless peripherals in a public space like a co-working space that might end up with other wireless devices nearby or physical access (say, at nighttime). My working assumption is that it would be reasonable to imagine that the keyboard could be attacked by some form of man-in-the middle but that this would be fairly sophisticated, e.g., not easily downloadable and able to be run on a single computer that happens to be nearby. Does anyone know of any projects checking this assumption & whether:
- It can be done without physical access and only in close wireless proximity
- It can be done, but, requires swapping the keyboard once the bluetooth connection is cracked to dual-transmit
- It can be done, but, would require first getting access to the computer [user login password] via some other means to change the remembered keyboard
Similarly, were there any security changes on a particular generation of magic peripherals or MacOS that are important? I assume that the Mac hardware itself is mostly independent except of course as far as exploiting the device for direct access to the computer which isn't the scope of this question.
Lastly, what checks are present when a keyboard is plugged in via lightning & what is the benefit in such a threat model to:
- Never re-connecting the device
- Reconnecting on arriving to the space before typing
- Only leaving the keyboard plugged in
It seems possible that even in the last model a keyboard could be plugged in and still transmitting or alternatively that realistic attacks might require #1.
