Maybe I'm missing something but how could finding out who's yodeling a movie be rather easy when you would have to decrypt the traffic to determine if it was a movie and not just normal traffic? I get that because of TCP/IP you can tell someone is using I2P but wouldn't you have to compromise the garlic encryption layer to determine what exactly they are doing?

[–] ninjan@lemmy.mildgrim.com 4 points 7 months ago

That's what I'm saying. It's like everyone knows some college kids smoke pot from the smell in the dorms, but Police can't legally search room by room to find out who it is, they need a search warrant which they need more than a general suspicion that someone in the dorms smoke to get.

Same with I2P, it's done in a public setting so from traffic patterns we can be pretty sure someone is downloading a shit ton, and that it's likely illegal content. Residential IPs have little reason to consistently download several GB files on a daily/weekly basis, streaming and download also look vastly different profile wise and at least no one I know of go to those lengths to try and mask their traffic patterns by trying to make streaming look like download or vice versa.

But as I said and you reiterated, you still need to crack the encryption to actually prove it in court. But given a specific target there are many ways to do that. A generic approach is likely not going to happen. Which means that I2P is secure much like having a secret chat in a crowded place like Grand Central Station in NY. You know that people are meeting there to chat about illegal stuff but you don't know who. It becomes much easier if you know who to follow and eavesdrop on, but of course still not easy.

It is however nowhere near as safe as communication over channels that aren't public to begin with. But such of course do not exist outside military and other special contexts.