In many ways, this is the same as Google/Amazon/Apple claiming they are not listening to voice commands 24/7, when anyone with common sense understands that they have to, to register the initial request too.
I don’t trust them to limit it to what they claim they’re doing
The distinction there is that they detect the activation command client-side and start sending to server after that. Chat control could technically be implemented in a similar way, where the client decides when and what to send to central processing. Only problem is… do we do all detection client-side then? Or do we let the user activate the detection by saying “Hey Interpol”?
In many ways, this is the same as Google/Amazon/Apple claiming they are not listening to voice commands 24/7, when anyone with common sense understands that they have to, to register the initial request too.
I don’t trust them to limit it to what they claim they’re doing
The distinction there is that they detect the activation command client-side and start sending to server after that. Chat control could technically be implemented in a similar way, where the client decides when and what to send to central processing. Only problem is… do we do all detection client-side then? Or do we let the user activate the detection by saying “Hey Interpol”?