SMAPs: Short Message Authentication Protocols (Transcript of Discussion)

Khaled Baqer: What I’d like to do first is to highlight the background and motivation for the payment project that we’re working on at Cambridge. I’ll do a ten-minute introduction and Ross will take over to discuss some of the attacks and the interesting parts of this paper. The story begins with the mobile payment revolution. This is not Apple Pay, Google Pay or whatever extension of EMV you have on your phone. These are mobile payments in LDCs, less developed countries. So top right, you have M-Pesa in Kenya, top left you have bKash in Bangladesh. These systems work by providing a menu for the user on the phone’s SIM toolkit (applet running on the SIM), so they can access, for example, ‘send money’, enter the amount and the recipient’s phone number. Usually this is saved in the user’s phonebook, so all you have to do is to enter the amount. The phone sends a message to the server, and the server replies with a confirmation back to the user, and the recipient gets a confirmation of the transaction. This has been transformative in bridging the gap and providing financial inclusion. Millions of people don’t have bank accounts, and even if they do, these accounts are dormant and nobody really uses them. Because of social reasons we can discuss offline, people don’t want to go to banks. Some of them don’t even know how to use their accounts, and they don’t want to walk miles to the closest bank branch. The first big use for M-Pesa was remittances: someone working in Nairobi sending money back home to their relatives in a remote village, and then they can cash out with the local agent. Store-of-value, of course, and personal safety, these are big ones: you don’t want to carry cash in some regions, it’s unsafe. It also provides a means for governments to deliver payments directly to users. Meaning if they want to have a direct financial relationship with the users they can directly send the funds to their phones and not send cash that can be lost along multiple hops, if it gets delivered to the intended recipient at all. What we started out with are the challenges listed here: can we extend this to areas where there is no network? Because mobile payments work beautifully if they work, but they stop when the network does, and so we have people in