In message "Clipper and Traffic Analysis", pmetzger@lehman.com writes:
Normally, one can only determine the endpoints of a conversation. With clipper, however, one can deduce a lot more, since when people move around, go to hotels, phone booths, etc, you can still track their clipper serial numbers.
Perry
Maybe we are missing something here, when people move around if they simply use their hotels phones, phone booths, et al, how can one track their clipper serial number? Unless there is a personal Clipper serial number? I do not see any extra info being garnered from Clipper phones, just less. Consider the case where there is a wiretap in progress: In the past one end was tapped and both sides of the converstion were heard. You could always find out who was calling, but you could not go and bug the calling party's phone without a court order. With Clipper, you tap the outbound voice/data, but every inbound voice has to be decoded with its own key. Now will the wiretap allows blanket decryption for all Clipper phones? and if so where does the tap begin and end. If no blanket decryption is allowed then must they record outgoing voice/data and based on the content of one side of the conversation convince a judge to let them decode the other side? Both scenario is problematic. If I'm cop and the suspect does not have a Clipper phone at home? can I choose which end of the switch to tap? I would choose the analog side that still has voice -- both voices, so even if the network is Clippered, why go through the hassle of managing keys when I can always hook onto an old fashioned analog voice line. Am I seeing correctly or did I miss something?