4 ms·

> Can a Kong manufacturer build custom secure element chips that leak key material? It seems that a lot of Kong's security comes from the security of the little

> Can a Kong manufacturer build custom secure element chips that leak key material? It seems that a lot of Kong's security comes from the security of the little HSM, and if you don't trust the manufacturer then a lot of it seems to break down. This is perhaps the most important question. In short, yes. We use an off the shelf secure element; if it's shown to be broken/backdoored it will influence a vast number of embedded and IoT applications. > In addition, the paper cites the cost of decapping and side channel equipment as part of what makes double spends impractical. Given that they're fixed costs that undoubtedly many governments and labs already own, it still seems like it could be worth it for some people to extract Kong keys. We cite fixed costs in hardware, but not in time. Certain extraction techniques would still take hours of time per chip. The secure element we're using has no published attacks thus far (which is not to of course imply that it won't be, simply that it isn't trivial).

eat_veggies 7y ago
> We use an off the shelf secure element; if it's shown to be broken/backdoored it will influence a vast number of embedded and IoT applications The scenario I'm imagining is one in which someone manufactures a note that looks and behaves exactly like a Kong but has known key material (i.e. a "fake" secure element). You can't verify that the secure element is real without expensive hardware. And what is the vision for who can mint Kongs? If only the Kong developers can, then it's no better than USD. If anyone can do it, how will you guarantee consistent chips, construction, design, etc. between manufacturers?
- ccamrobertson 7y ago
  > The scenario I'm imagining is one in which someone manufactures a note that looks and behaves exactly like a Kong but has known key material (i.e. a "fake" secure element). You can't verify that the secure element is real without expensive hardware. Yes, that's counterfeiting in the case of Kong. The counterfeiter would need to first extract that key material from a previously issued Kong note by breaking the secure element. > And what is the vision for who can mint Kongs? If only the Kong developers can, then it's no better than USD. Only the Kong project for now; whether it's better or worse is definitely subjective. There is a ceiling for the amount of Kong that can be issued. Also, as something more akin to an art project than a fully functioning economy, there is no place to spend Kong today. It's ambiguous, but the fact that the Kong project is the only issuer of the physical notes doesn't undermine the cryptocurrency rules upon which it's based. > If anyone can do it, how will you guarantee consistent chips, construction, design, etc. between manufacturers? Section 5.1 touches on this; in short there is no solution today, but this in an immensely important challenge. Flipping this on its head -- if we don't have these guarantees today, how can we trust any of the places we store key material (smartphones, laptops, IoT). It's impossible to guarantee a perfect chip, but it is feasible to create economic incentives that deter broken chips. We'll probably write more on this topic in the future.

eat_veggies 7y ago

> We use an off the shelf secure element; if it's shown to be broken/backdoored it will influence a vast number of embedded and IoT applications The scenario I'm imagining is one in which someone manufactures a note that looks and behaves exactly like a Kong but has known key material (i.e. a "fake" secure element). You can't verify that the secure element is real without expensive hardware. And what is the vision for who can mint Kongs? If only the Kong developers can, then it's no better than USD. If anyone can do it, how will you guarantee consistent chips, construction, design, etc. between manufacturers?

ccamrobertson 7y ago
> The scenario I'm imagining is one in which someone manufactures a note that looks and behaves exactly like a Kong but has known key material (i.e. a "fake" secure element). You can't verify that the secure element is real without expensive hardware. Yes, that's counterfeiting in the case of Kong. The counterfeiter would need to first extract that key material from a previously issued Kong note by breaking the secure element. > And what is the vision for who can mint Kongs? If only the Kong developers can, then it's no better than USD. Only the Kong project for now; whether it's better or worse is definitely subjective. There is a ceiling for the amount of Kong that can be issued. Also, as something more akin to an art project than a fully functioning economy, there is no place to spend Kong today. It's ambiguous, but the fact that the Kong project is the only issuer of the physical notes doesn't undermine the cryptocurrency rules upon which it's based. > If anyone can do it, how will you guarantee consistent chips, construction, design, etc. between manufacturers? Section 5.1 touches on this; in short there is no solution today, but this in an immensely important challenge. Flipping this on its head -- if we don't have these guarantees today, how can we trust any of the places we store key material (smartphones, laptops, IoT). It's impossible to guarantee a perfect chip, but it is feasible to create economic incentives that deter broken chips. We'll probably write more on this topic in the future.