Proxies

How useful would a proxy (perhaps running on a VPS) only used by a single person be, in trying to hide from a very advanced threat actor (like the NSA)? Assuming the NSA has complete access to all of the ISPs logs, and can pretty much see all Internet traffic, surely they would be able to see that you're making a connection to the proxy server from your home IP address, and be able to tell that that proxy is just relaying all your data. Then. they can look at what the proxy is sending or receiving. This would make using a proxy to hide your activity from a threat actor like this practically pointless, right? Does this mean that the only useful proxies are those used by others, that let you blend in with other users, or am I missing something? Does a proxy chain help at all if you're still the only one using the proxies?

>>23374 >How useful would a proxy (perhaps running on a VPS) only used by a single person be, For downloading entirely legal Linux ISOs? Might be useful. >in trying to hide from a very advanced threat actor (like the NSA)? Utterly useless and unrealistic. What is your threat model? All these questions about how to hide from NSA or some other nation-state entity always neglect to consider whether they're actually at risk from them. If the NSA is interested in you *personally* then you likely already screwed up/"lost" by posting about it. But just to play along: what you need is separate identities, e.g. one for posting on public forums and another for laundering data from NSA servers. Determine an appropriate threat model and mitigations for each identity and don't let information linked to one identity cross over to another identity. If you're stealing data from the NSA, consider using microSD cards and carrier pigeons.

>use pigeons based And you're right, if an entity or organization with virtually infinite resources, time and lack of legal accountability like the NSA has targeted you in particular then it's a miracle if you get to post anything here at all. This is not to be confused with assuming them to be omnipresent in some way but they are scary. But seeing that OP is not only a user but also (still?) alive, this does not seem to be the case. The advice about strictly separating identities (they got the pirate bay guy on a single use of an old email from a dusty forum post) is a sound one. However I do wonder how the ability of a person to not accidentally stick out could be improved. For instance would it help if everyone were to switch to i2p over night? OP's idea might be bunk but could there be an approach to networking systems that make it impractical to shut down even for an overbearing government unless they go full authoritarian? And even then some mainland chinese are still trading crypto, which officially is illegal there iirc.

>>23376 Yeah, I’m aware that if you’re being targeted there isn’t much you can do. I’m more thinking about how we can, as a society, get to the point where no one, even those who control the Internet infrastructure, can deanonymize large numbers of us. Should’ve made that clear. That doesn’t really matter though. My question was specifically about proxies, and whether they actually provide any kind of anonymity from these kinds of threat actors if your traffic is not being mixed with other people’s.

>>23380 Anon from before. I hope >>23376 comes back to elaborate more but essentially just think about on a technical level what it is that you're asking: Does a proxy substitute your IP address with its own by routing all your traffic through its connection thus making any requests look like they originate from there? Yes, that's basically what a proxy, VPN, technically your ISP is. Do authorities inquire that service to whom they've lend the usage of their IP address when they want to know who it is that made a particular request which caught their attention? Be it the host of your privately run proxy? And they willingly hand over anything they have of you? In the overwhelming amount of cases, absolutely! So in a direct way it does provide anonymity which is good if you want to evade ad trackers and watch region blocked content but down the line if you run your secrit amonumous haxxor account on x formally known as twitter then eventually a couple suits will knock on your home door. There's ways to greatly mitigate this by using traffic routing schemes with the explicit aim to ensure anonymity like tor and i2p and they do work quite well seeing as it takes tremendous amounts of resources and cooperation of multiple public and private entities to pin down a single onion site but they also come with a long bucket list of do's and don'ts where a single slip up get's you busted. In addition adoption is a bit lack luster too, which means little legitimate services using it besides the usual dark web creeps. Also tor has a scalability problem and probably only works as well as it does due to half the nodes being operated by feds. It means, I'm afraid, there not really being much right now that you can do and a lot more has yet to be built. But I hope the urgency of that situation will compel enough people with a mind for that sort of stuff to spend some of their time and creative power on it. Maybe you're one of those.

>Does this mean that the only useful proxies are those used by others, that let you blend in with other users, or am I missing something? Yes, you are correct. The only worthwhile anonymizers are onion networks, because of the fact that they permit you to mix your traffic into a network comprised of any participants within the network whom can be found around the world. Traffic originating from the network could be the product of any one participant. Any decent onion network will itself not make efforts to dissuade individuals from hosting nodes for the network, allowing for the potential of a large portion of individuals to host nodes, and a much larger amount of people to use the network. In theory, an onion network has the potential to be trustless, in that in order to safely use the network one is not necessarily required to have complete trust in the entirety of the nodes comprising the network to not be malicious. A number of nodes can be maliciously used by bad actors to exploit users of onion networks, but for these attacks to have any meaningful prevalence in the network would require a massive expenditure of resources for the sake of hosting malicious nodes to dominate the network. If actions aren't taken to dominate the network, the malicious nodes should remain in the minority granting that average people go out of their way to host nodes. So long as the users of an onion network have good practices, like disabling javascript in their web browser, and never downloading content with an unencrypted connection through the network, the threat of malware spreading and information passing through the network being stolen by criminals can be almost completely eliminated. The most prominent remaining threat would be that of a passive deanonymization of onion network users by a governmental entity seeking to monitor their activities. The most practicable method of deanonymization would be by the way of timing attacks. In the case of timing attacks control over the first and last node in a victim's circuit through an onion network is required by the perpetrator. When considering the fact that theoretically just about anybody can host nodes for an onion network, the efficacy of timing attacks in their ability to be executed with frequency appears severely hampered, as a perpetrator's nodes should be in the minority, and their ability to be executed with enough precision to target specific individuals appears near non-existent. So long as an attacker doesn't have the ability to manipulate the manner in which a user's client for their onion network selects nodes to connect to, the attacker cannot have any more autonomy over the selection of victims than randomness grants them, in which case increasing their number of nodes is their only viable option, for the most part. A perpetrator could easily spin up nodes serving to passively deanonymize users of the network, which would over the passage of time allow the perpetrator to correlate the activities of network users' sensitive-activities-intended-to-be-kept-private to their real identities. Only a government like that of the United States would have any desire to run such an operation, and the United States government is certainly running one. Furthermore, the U.S government has the capital to host a large amount of nodes, maybe enough to dominate a network like Tor. The Tor network is designed to be a low-latency onion network, making compromises between the security of users' anonymity and network speeds to try and find a reasonable balance between speed and security. By making these compromises they greatly increase the marketability of the network. An onion network with high latency, while indubitably more resilient to timing attacks resulting from the attributes of the network that result in its high latency, those attributes for example being a time delay of a random duration in the relaying of traffic through nodes across the network, will almost certainly not be seen to have much utility by an average person due to its extremely sluggish download speeds. And so you'll have far less people using the network than you otherwise could by having faster network speeds, less people to hide amongst, weaker anonymity. Part of the way the Tor network maintains a low-latency for users is by instilling in users' clients a predilection for low latency nodes within the network. This of course opens the door for exploitation. An entity seeking to increase the efficacy of their timing attacks against users of the Tor network needn't do more than ensure that their nodes are running at a lower latency than that of other nodes. For an entity like the U.S government with nearly infinite resources, the hosting of low latency nodes should not bring them great strain. Now I rambled about onion networks largely from a base level, considering the concept of them, although I am now diverging more into the discussion of actual onion networks I guess, of which there are only three major networks, those being Tor, i2p, and lokinet respectively. They all have their differences. i2p for one is not intended to be used to interface with the remainder of the internet, instead designed to function as "its own internet" so to speak, in essence being an autonomous network built upon the internet, providing great anonymity and freedom to its users. Lokinet has a unique methodology which attempts to disincentivize sybil attacks. I sort of mentioned sybil attacks earlier. For a sybil attack in the case of an onion network, an attacker would subvert the network's reputation system using numerous separate identities under their control so as to attain a disproportionately large influence over the network, to the point they could severely lower the prevalence of non-malicious nodes, from whence they could more effectively execute timing attacks, for example. It's basically the means to an end, the part of a greater attack, for if a monolithic entity appeared emerged within an onion network, controlling a disproportionately high portion of its nodes, it would in all likelihood be eliminated out of suspicion before being asked to relinquish some of its power. Lokinet attempts to disincentivize sybil attacks by forcing nodes to conduct mathematical algorithms by the way of Proof of Work in order to maintain their existence within the network, a concept born out of Bitcoin, I don't even wanna get into all the clusterfuckery of lokinet with oxen and their technologies, the nodes literally mine crypto and profit is generated by it, but it is costly, a lot of computing power is required, and so it costs real money to run a lokinet node. Now, you can completely destroy a criminal's ability to execute sybil attacks with this methodology, but not so much a government. I dislike lokinet for this. As I said it costs real money to host nodes -unless things have changed from what I last read. The vast majority of people aren't going to want to host nodes because of that. Why, the vast majority of people don't even want to host Tor nodes and it costs much less money, if they don't want to host Tor nodes out of cost then it's simply out of convenience, for most the thought of hosting a node has likely never crossed their mind, they have never tried to host a node, they've never been forced by the police or their ISP to stop. As the wealth gap between the rich elites and us, the normal people, grows, and we find ourselves struggling more and more to pay the bills, sinking deeper into poverty, tearing each other apart bit by bit in this rat race, do you really think that people are gonna be able to spend hundreds to thousands of dollars on running a damn lokinet node? No. And do you think the government which we have to pay tribute to, offering it trillions of dollars, the government with its various alternatives streams of revenue born out of warfare and crime in which profits are funneled back toward them through taxes, do you really think they're not going to have the money to run nodes? Or the major corporations working alongside them? And does it matter in the grand scheme of things that people won't run nodes, allowing a government to supplant the network? That depends, it could matter, it would be in the service of the death of digital freedom, a service seemingly trying to provide that freedom would become corrupted. I believe it ultimately to be corrupted from the start, particularly in how profitable it must be for the developers to run. An onion network that truly intends to provide users with freedom by the way of anonymity and privacy that is based upon money, and is directed towards profit, is doomed to fail. I don't think the lokinet devs even intend for it to be a truly libre software. To be able to use something in a truly anonymous nature, you cannot have control over that thing. If you control that thing, it follows that that thing can be linked to your identity in some way, deanonymizing you. You can allow other people to use your thing, and try to mask the business you conduct with your thing among the group of people using your thing, but still, you own the thing. It is kind of the logical way of things. If you own something, it can be definitively associated to your identity, if you don't, it can be associated your identity but not definitively. You can use a VPS, a virtual private server, essentially just a server you rent and can control digitally and use as a proxy between websites. Now, if you just used this server alone, there'd be nobody to hide amongst. If someone did something that ruffled the government's feathers so to speak, and they decided they wanted to track this person down, and they traced this person back to a VPS, I think it'd be safe to say that they'd have a pretty easy time identifying their person of interest.

Something to note in hindsight, as I stumbled into this problem with the following paragraph. The clusterfuckery of proxy definition by tech illiterate people (and I'm not even tech literate). A VPN is just a server you do not own that you proxy your internet traffic through, okay? Call it a proxy/proxy server, VPN server if you want, whatever. It's a proxy you pay for, provided by a company running a number of other proxies, that is it. You connect to it using an internet protocol designed to ensure a secure, encrypted connection to the (VPN's) proxy server. A VPS is just a server provided by a company that provides servers for you to remotely connect to, and use for the purposes permitted by the company. The difference is you actually have (digital) control over the server. A SOCKS5 proxy provided by a company, is that a VPN? No. Why not? Because a VPN server has to fall under some other protocol? Why? VPNs are just overglorified proxies, as someone else said. A VPN and VPS can be misconstrued as being the same thing over the protocols used to interface with them, the protocols are the same for the most part. Now, you could proxy into Tor through this VPS. I believe this to be gratuitous for those living in a free country where Tor is legal. And for those living in countries banning Tor, well, I don't think what they do even matters as most VPN protocols are blocked in their countries. VPNs are just as easily derezzed as Tor, there is merely less stigma behind VPN usage since they're ineffective. In countries hostile towards anonymizers the ISPs usually block at a protocol level as opposed to blocking specific servers. What point is there in using a proxy before Tor? Hiding Tor usage from your ISP, which is controlled by your government? Who's to say that your proxy doesn't have some backroom deal with your government, feeding them logs? I guess it depends on the kind of proxy. For a VPN, it's a certainty. For a VPS, it's not a certainty, but it very well could be happening, and there's no way for you to know since you don't physically control the VPS. If you came under investigation, it'd certainly happen, the company providing you the server isn't any less obligated to abet the authorities than a VPN company. Any of the deanonymizing attacks where your activity in real life would be correlated with Tor activity would not be rendered inactionable by the mere usage of a VPS to mask a connection to Tor, nor would they be made any more difficult to execute. To me, the people using proxies before Tor needn't do more than cross their fingers, and hope. Hoping for what? That is just works out, that "they" don't find out they're (proxy->Tor users) using Tor. So fuarrrking what, people should be using Tor for a free internet, and a free world, and people's fear of landing on a government watchlist over trying to not have Big Tech 30 feet up their ass is a problem fueling our technological enslavement. I for one won't take ineffective measures to hide my usage of Tor. I believe more people should be using Tor, for two, and if I willfully decide to use it regardless of whatever consequences may be brought upon me in favor of my reasoning for using it then I may inspire others to do likewise, this I view to be more valuable than trying to hide away ineffectively. As for what I'm doing on the network is business being mine. You absolutely cannot be the only one using the proxy. So you need to be using a proxy that is free to all, one that anybody anywhere is allowed to use, otherwise you'll have to use a network requiring account registration, and your anonymity is gone. More people need to use onion networks, namely Tor and i2p. In order to bolster their efficacy, people need to be willing to host nodes. There are issues with hosting exit nodes, not so much for hosting entry nodes though but I think the Tor Project makes hosting entry nodes a little too inaccessible to people.

>>23379 >For instance would it help if everyone were to switch to i2p over night? If you could convince everyone to switch over to I2P then you could also convince everyone to stop allowing the existence of organizations like the NSA. Anonymizing solutions only treat the symptom and only for some time if evil stays unopposed, since they'll vice down harder and harder if left unchecked. >>23380 >>23381 >and whether they actually provide any kind of anonymity from these kinds of threat actors if your traffic is not being mixed with other people’s. For personal proxies the short answer is no. Maybe if you get some servers outside five-eyes influence it'll be harder to monitor traffic but they can still see you're among the very few using a personal server in a foreign country. With Tor/I2P (or to a lesser extent popular proxies such as Mullvad) your traffic gets mixed in with others, so it becomes harder to correlate traffic. >>23383 >>23384 Excellent posts.

>>23387 As usual, the answer to 1984 is 1776. But in the meantime I'd like to know whether we can have an unobstuctable and unredactable internet if just enough people got on board with it.

>>23374 Tor initially evolved out of the idea that plain proxies are insufficient for keeping your IP address hidden. When you use something like a socks5 proxy (I don't know about other protocols), you have to tell the first hop where to go for the second, the third, and so on. The first hop knows all of your hops. Tor solves this by negotiating encryption for every hop, so every hop only knows the previous and next hop. This is the eponymous onion routing in The Onion Router, and is why there are three hops for a typical Tor route (the middle node doesn't know the source nor the destination, breaking the chain of knowledge). Other improvements came later, like only sending uniform-length packets so adversaries can't decloak you by measuring your specific sequence of packet lengths. And other measures; there are a lot of them.

>>23384 >A SOCKS5 proxy provided by a company, is that a VPN? No. Why not? The problem here is in defining a VPN as a proxy. If VPN = proxy, and SOCKS5 = proxy, then SOCKS5 = VPN. But in fact, the essential definition of a VPN is not in its application as a proxy. This is one use of a VPN, but not the main purpose or essence of it. The clue is in the name, virtual private network. Note that the "VPN" is not the same thing as the "VPN gateway" or "VPN server". You start by talking to a VPN gateway, but after that, the result is not simply that you're now routing all of your traffic through that VPN gateway. The real point is that your computer now has a virtual network interface with an RFC1918 address that is in your VPN's "encryption domain"; i.e., the LAN or portion of the LAN on the remote end that houses the resources you need to access without actually being Local to it. Now you can be wherever, and you can access your company's internal resources as if you were in the office with them, without worrying about NAT (which is, of course, materially different from encapsulation even if the effect is the same) or exposing the internal resources to WAN (giving them an external IP address and making them accessible directly from the public internet). Or you can have two office buildings in separate places that can operate as one LAN using normal internet connections instead of paying for a leased line, for example. Now you are Virtually on the same Private Network. The facts that the traffic is encrypted in the tunnel and that your apparent origin IP is now the VPN gateway are merely side effects of how VPN works, not the actual point.

>>23383 A correction to make to my post (I knew this but didn't say so for some reason) an adversary only needs to control the entry node and website for a timing attack to be successful. I'd go further with that and assume that if they had access to your router (which can be obtained via TR69 backdoor if present) then the entry node could become unnecessary. This is highly troubling consider most web traffic is centralized to the Big Tech conglomerate.

your best option is hiding a raspberry pi at a random house on their 2.4ghz wifi and using it as a vps you only connect to through tor. maybe.

>>24078 ppl usually say this and go tor add-on but whats the point if already vps-ed through raspberry? services like oxen or torsocks sometimes soften up vulnerability to timing attacks: whats the best torrificator in combination with raspberry? tails/ whonix oracle oxen etc

>How useful would a proxy (perhaps running on a VPS) only used by a single person be, in trying to hide from a very advanced threat actor (like the NSA)? anything like the NSA? virtually useless. If your threat actor in question is a intel agency you need several layers of defense. >Does this mean that the only useful proxies are those used by others, that let you blend in with other users, or am I missing something? Does a proxy chain help at all if you're still the only one using the proxies? Depends on your theat model, proxies are good for hiding to the end hosts and man in the middle. Several people doesn't mean anything because the vpn itself can correlate who is doing what using their vpn. Most of the time they're gonna reach out to the vpn or some other de-anonymizing techniques, like fingerprinting and time correlation.