I am not sure what he’s hinting at. Just using Tor doesn’t bear any legal risks. Hosting an exit node is different, as depending on the country you might get into serious trouble if certain traffic goes through it.
Would it be possible to allow exit nodes to blacklist specific kinds of traffic and somehow privately verify that the traffic is not one of the blacklisted kinds (zero knowledge proof perhaps sorry not a CS person)?
An exit node can put in place any filters, blacklists, mitm, exploit injection, logging, and anything else it wants… on unencrypted traffic. Using HTTPS through an exit node, limits all of that to the destination of the traffic, there is no way to get a ZK proof of all the kinds of possible traffic and contents that can exist.
Yes exactly, and I think there have been stories recently where the exit node host has been held liable for content that’s gone through it.Which is complete bullshit, but the unfortunate reality is that the legal system doesn’t need to understand technology to regulate it.
It’s not bullshit. If A has proof your system launched an attack, or sent CSAM, to another system, but your only defense is “I let anyone use my system in that way”, then at the very least you’re an accomplice.
It is bullshit, because it puts the onus of policing everything on any service provider. If a TOR exit node provider is responsible for all traffic through their node, then an ISP is responsible for all traffic through them to their users - yet it is not reasonable for ISP’s to do this. Nor should it be acceptable by law and even less so if the purpose is for law enforcement to bypass the warrant system by having private parties do the investigation for them.
Well, the law enforcement ship has sailed a long time ago, it’s more of a flotilla by now. Data communication service providers (including ISPs) have some customer identification and data retention requirements in exchange for immunity from the data itself, but otherwise —reasonabke or not— there are more and more traffic policing laws that get introduced for ISPs to abide. By starting a Tor Exit node, you become a service provider, and the same laws start to apply.
It’s no joke that we live in a surveillance state, just that some go “full surveillance” like China, while others go “slightly less in-your-face surveillance” like the US/EU.
To give you an idea, last time I used Tor, I suddenly started to get a bunch of connection attempts from the FBI. Was I doing anything illegal? Nope. Was TOR a legal liability? You betcha.
I was using peerblock and one of the blocklists contained known governmental IP addresses. Those blocked connections began quickly filling the logs.
Spooked the crap outta me. It’s been a few years since I did that, so I could have that detail wrong. I know it was for sure one of the three letter acronyms, DOD, FBI, CIA, but they were definitely incoming.
That does not sound plausible to me. Typically, your own computer would be behind a router that is either doing NAT or has a firewall (probably the former). Any incoming traffic would be directed to the router without any chance of reaching your computer. Whatever you saw was either outgoing traffic or incoming traffic in response to connections initiated by your own computer.
Consider this, the Tor software was accepting connections from government IPs.
Regardless of whether it was active intrusion or a significant portion of the Tor network, (at that time) had a number of governmental IP ranges in it, It’s enough to dissuade my use, at least without more significant OpSec.
I use peerblock and had some good blocklists set up. The hardest part should be finding peerblock or a more modern fork, the blocklists are mostly public. Helps keep from connecting to known bad actors.
Which are what
I am not sure what he’s hinting at. Just using Tor doesn’t bear any legal risks. Hosting an exit node is different, as depending on the country you might get into serious trouble if certain traffic goes through it.
Would it be possible to allow exit nodes to blacklist specific kinds of traffic and somehow privately verify that the traffic is not one of the blacklisted kinds (zero knowledge proof perhaps sorry not a CS person)?
An exit node can put in place any filters, blacklists, mitm, exploit injection, logging, and anything else it wants… on unencrypted traffic. Using HTTPS through an exit node, limits all of that to the destination of the traffic, there is no way to get a ZK proof of all the kinds of possible traffic and contents that can exist.
What I meant was blacklisting certain destinations. It obviously wouldn’t prevent all malicious traffic
Yes exactly, and I think there have been stories recently where the exit node host has been held liable for content that’s gone through it.Which is complete bullshit, but the unfortunate reality is that the legal system doesn’t need to understand technology to regulate it.
It’s not bullshit. If A has proof your system launched an attack, or sent CSAM, to another system, but your only defense is “I let anyone use my system in that way”, then at the very least you’re an accomplice.
It is bullshit, because it puts the onus of policing everything on any service provider. If a TOR exit node provider is responsible for all traffic through their node, then an ISP is responsible for all traffic through them to their users - yet it is not reasonable for ISP’s to do this. Nor should it be acceptable by law and even less so if the purpose is for law enforcement to bypass the warrant system by having private parties do the investigation for them.
Well, the law enforcement ship has sailed a long time ago, it’s more of a flotilla by now. Data communication service providers (including ISPs) have some customer identification and data retention requirements in exchange for immunity from the data itself, but otherwise —reasonabke or not— there are more and more traffic policing laws that get introduced for ISPs to abide. By starting a Tor Exit node, you become a service provider, and the same laws start to apply.
It’s no joke that we live in a surveillance state, just that some go “full surveillance” like China, while others go “slightly less in-your-face surveillance” like the US/EU.
To give you an idea, last time I used Tor, I suddenly started to get a bunch of connection attempts from the FBI. Was I doing anything illegal? Nope. Was TOR a legal liability? You betcha.
Connection attempts from the FBI? Could you specify that a bit further?
I was using peerblock and one of the blocklists contained known governmental IP addresses. Those blocked connections began quickly filling the logs.
Spooked the crap outta me. It’s been a few years since I did that, so I could have that detail wrong. I know it was for sure one of the three letter acronyms, DOD, FBI, CIA, but they were definitely incoming.
That does not sound plausible to me. Typically, your own computer would be behind a router that is either doing NAT or has a firewall (probably the former). Any incoming traffic would be directed to the router without any chance of reaching your computer. Whatever you saw was either outgoing traffic or incoming traffic in response to connections initiated by your own computer.
Consider this, the Tor software was accepting connections from government IPs.
Regardless of whether it was active intrusion or a significant portion of the Tor network, (at that time) had a number of governmental IP ranges in it, It’s enough to dissuade my use, at least without more significant OpSec.
I do understand your point though.
How can I observe connection attempts like this?
I use peerblock and had some good blocklists set up. The hardest part should be finding peerblock or a more modern fork, the blocklists are mostly public. Helps keep from connecting to known bad actors.
Yeah, is this guy living in China?