[daeken]

> How do you plan to reconcile the web of trust with the potential harm that can come to people who leak data?

So, the web of trust doesn't apply to individuals, just nodes in the network. Thus, each node can handle incoming leaks however they want to, or not at all. In fact, if this is designed properly then there might be some sort of plausible deniability in terms of who got a leak first.

> I think the real issue is verifying the legitimacy of a leak. If you can verify something, you don't have to trust the person who posted it and they can remain safe and anonymous.

Verifying the legitimacy of leaks is more of a social than a technical issue; the web of trust is there primarily to prevent propagation of damage, not to ensure propagation of the truth.

> Also, are you familiar with http://globaleaks.org/ ?

Yeah, it's a nice project, but the lack of a federated protocol makes it vulnerable to the same thing as Wikileaks and friends.