> But Freenet is still around, although it's too risky to use it, except via Tor.
Hi, I'm a Freenet developer of the past ~ 10 years, so I'd like to clarify upon this :)
(The project is still active, there was a release just this week!)
While there technically were indeed lawsuits in the US, the situation is not as black and white as "it's dangerous".
It is an anonymizing peer-to-peer network as well!
What is dangerous under certain circumstances is only one of the three modes to use it:
1) Opennet, where Freenet uses random strangers as peers.
2) Darknet, where you only connect to peers you manually select, e.g. your friends.
3) Opennet with some Darknet peers in addition (I'll call it "mixed mode").
So Opennet allows law enforcement to connect to your Freenet potentially and thus analyze your traffic.
Still, this does not mean that your Freenet will plainly tell its peers what you are downloading!
Traffic is always redirected across a random number of peers, none of which tells the others who requested it - which provides plausible deniability.
All traffic is encrypted, only the recipient can decrypt it.
So you cannot just watch traffic and filter out illegal JPEGs or whatever.
What LEA did then is to come up some math and then claim to deduct from it that there is a certain probability that the illegal downloads were requested by the people they claim it came from.
Their math is known and discussed by the Freenet core team, it may be addressed eventually - but from watching the discussion (not the math) I can say it should be taken with a grain of salt.
It's not absolute proof that the claimed downloaders were in fact the downloaders.
It's just a probabilistic assumption, which may possibly be wrong because the way Freenet works is rather complex (>200 000 LOC).
So as Freenet stores content encrypted on random user's machines (which is the advantage over Tor, Freenet is completely decentralized!), it is imaginable that law enforcement accusses people who did not willingly download it, but just happened to store it.
But: You can use Freenet in Darknet or mixed mode to be reasonably safe:
The more of your peers are not controlled by attackers, the lower the probability that a statistical attack can be conducted.
Further, the said legal cases only happened in the US to my knowledge, and I'd argue that the legal system of that country seems a bit flawed.
Outside of the US you can just run Opennet and probably be at the same risk as some random non-exit Tor node.
You transport traffic which you cannot look into (because its encrypted) and store files which you cannot look into (because they are encrypted), so what's illegal about it anyway?
> So Opennet allows law enforcement to connect to your Freenet potentially and thus analyze your traffic.
Further, it should be clarified that this is not a problem specific to Freenet:
ANY network which tries to be anonymous will suffer from the so-called "sybil" attack if it connects to random strangers:
If an attacker runs e.g. 100 000 machines on a network of only 1000 actual users then the probability that a single user only has connections to them is very high.
And anonymization must rely upon redirecting traffic across multiple peers - but it cannot if all peers belong to the attacker.
To my understanding Tor addresses this problem by heuristics, e.g. closely monitoring important, big machines in their network, trying to ensure they are in fact distinct entities - but that is really just guesswork, not hard mathematical security.
If Tor wanted to be truly secure it would have to add a darknet mode as well.
I should have been clear that I was talking about opennet mode. If you want to use Freenet in darknet mode, among people who know each other well, and trust each other, it's at least safer than (say) using a private torrent tracker. I mean, torrent traffic is also encrypted, these days.
It's true that you're relatively safe from adversaries, if you only use darknet mode. But there's always the possibility that one or more of your peers will get busted through some other exploit. And that they cooperate, and become informants.
But in darknet mode, you can only communicate with your peers, and can only access stuff that you and they have uploaded. If you want to communicate with the global opennet, and share stuff with it, at least one of your peers must have opennet peers. And that exposes them, at least, to adversaries.
If they get busted, and cooperate, others in the darknet are now at risk, because an adversary could use their client to probe its peers. They couldn't add other peers to the darknet, however, without some social engineering.
So anyway, it's whatever nodes that peer with the global opennet which are the main risk. And to do that safely, one can use anonymously leased throwaway VPS as gateways to the global opennet. You reach them via Tor. So if they go down, adversaries don't learn anything actionable about the darknet itself.
Hi, I'm a Freenet developer of the past ~ 10 years, so I'd like to clarify upon this :) (The project is still active, there was a release just this week!)
While there technically were indeed lawsuits in the US, the situation is not as black and white as "it's dangerous".
It is an anonymizing peer-to-peer network as well! What is dangerous under certain circumstances is only one of the three modes to use it: 1) Opennet, where Freenet uses random strangers as peers. 2) Darknet, where you only connect to peers you manually select, e.g. your friends. 3) Opennet with some Darknet peers in addition (I'll call it "mixed mode").
So Opennet allows law enforcement to connect to your Freenet potentially and thus analyze your traffic. Still, this does not mean that your Freenet will plainly tell its peers what you are downloading! Traffic is always redirected across a random number of peers, none of which tells the others who requested it - which provides plausible deniability. All traffic is encrypted, only the recipient can decrypt it. So you cannot just watch traffic and filter out illegal JPEGs or whatever.
What LEA did then is to come up some math and then claim to deduct from it that there is a certain probability that the illegal downloads were requested by the people they claim it came from. Their math is known and discussed by the Freenet core team, it may be addressed eventually - but from watching the discussion (not the math) I can say it should be taken with a grain of salt. It's not absolute proof that the claimed downloaders were in fact the downloaders. It's just a probabilistic assumption, which may possibly be wrong because the way Freenet works is rather complex (>200 000 LOC).
So as Freenet stores content encrypted on random user's machines (which is the advantage over Tor, Freenet is completely decentralized!), it is imaginable that law enforcement accusses people who did not willingly download it, but just happened to store it.
But: You can use Freenet in Darknet or mixed mode to be reasonably safe: The more of your peers are not controlled by attackers, the lower the probability that a statistical attack can be conducted.
Further, the said legal cases only happened in the US to my knowledge, and I'd argue that the legal system of that country seems a bit flawed. Outside of the US you can just run Opennet and probably be at the same risk as some random non-exit Tor node. You transport traffic which you cannot look into (because its encrypted) and store files which you cannot look into (because they are encrypted), so what's illegal about it anyway?