The first thing that I would like to observe is that a lot of the value in Internet is in its global nature. (a) Whether you think about your personal or professional lives, you are going to have friends and partners in different places. When that happens, localization isn't that useful. (b) Metcalfe's law says that the value of the network goes up when it has more participants. (c) You benefit from global innovation So from the outset, localization does not look such a great idea. Network localization can be good for some things, though - ixps, cables, cloud location - shorter rtt - resiliency - competition But not necessarily a solution to all problems - global aspects - network does not follow borders - do you guarantee that data does not flow anyway? a lot of intelligence operations are run on foreign land, malware does not look at borders Advocate thinking in terms of: - awareness of where your information is - trusting a limited number of parties to have your information - trustling a limited number of platforms for your devices - encryption technology for communication Examples - E-mail routing that forces messages to stay in country vs. using e-mail servers that encrypt your messages as they transit - Still have to rely on servers, but at least not on all routers in between - Too easy to corrupt or force just a single entity somewhere to leak information - Lets the trusted servers very carefully, however! Bad ideas - closed services for only the country - forced routing or data location Good ideas - cables, IXPs - information about where your data is - selecting trustworthy e-mail and other providers - diversity in what cloud and other services there are - competition in not just in price, but also in privacy and other things offered - encryption to cut down on the number of parties with your information