The title of this should be "Don't expect VPN to magically protect your privacy," not "Don't use VPN services."

Here are some reasons I've used, and continue to use, VPN:

* When I am on a network that uses an idiotic blacklist to block certain types of content. The network might even be run by my employer and I might be accessing content that is necessary for my work, but there might be no way to appeal the idiotic blacklist.

* When I am on a network that INJECTS content into HTTP responses (a certain paid airline WIFI used to do this).

* When I am on a network that might allow other users on the network to snoop on / mess with my traffic.

* When I want to access services that I have paid to access but are only available to IP addresses in a specific geographic region, and I happen to be in another geographic region.

Etc.

I used to be employed at a place that was so restrictive I couldn't even access asp.net (the website). I think it was something to do with it being in the cloud and looking like it was being hosted in the middle east. Most people probably don't know what it's like to work in a company with the extremely power hungry network admin that want someone coming to them for everything.

Three of your four points are explicitly addressed in there as reasons to use a vpn.

That github note doesn't really disagree with the article, which points out that you need to trust your VPN provider.

My general position is this: I don't trust my phone provider. At all. Just a week or so ago there was an HN post demonstrating how an ad provider can get your full name, cellphone plan details etc just by calling an API from a page rendered on your phone. But I also don't really have a choice - AT&T or Verizon or T-Mobile, they're all different flavors of the same crap.

Do I trust my VPN provider unequivocally? No. But I trust them a hell of a lot more than my phone provider, and they can't sell my personal info against my browsing history because they don't have it.

A VPN isn't the answer to everything, but nor is it useless.

Why do you trust your VPN provider more than your phone carrier?

What have they done to earn your trust?

It's a bit like how "stranger danger" isn't a thing kids get taught about anymore, because random strangers aren't risky if you go up to them, only if they come up to you. (Or, in more statistical terms: bad actors are a small proportion of the pool, but they have an incentive to self-select into interacting with you that good actors do not. If you just draw randomly from the pool, you won't get a bad actor. If you let the pool show the initiative, you'll get mostly bad actors.)

Your VPN provider is just some random company. You went up to them. They're randomly selected (insofar as your choices are random) from the space of all VPN providers, and most providers aren't malicious.

Your ISP is, at least in the US, almost always a monopoly. They're self-selected: they went up to you.

A VPN provider can tell you they're not logging your traffic because they think they aren't but really they are because there's a box somewhere that your traffic passes through that has logging enabled (for example -- and don't hyperfocus on the example, I know how you programmer types like to pick up the example and play ping pong with it for six hours).

So incompetence is a reason to not trust a provider as well.

Partially, at least, they don't need to earn my trust as much. They don't have my name, address, date of birth and social security number/credit data, like my phone company does.

The only positive point of trust a VPN provider has is that no-one has exposed them selling browsing data. Definitely not great, but also better than my phone company by default.

I suspect that in general there are two reasons.

* My VPN provider explicitly states that they do not collect user information or store logs of user activity. Unlike my ISP that has a No Privacy Policy.

* My VPN provider has not done anything to lose that trust.

So which is worse, your VPN provider telling you that they don't store logs of user activity and then very well doing it (as has been proven in multiple cases), or your cell provider telling you they're going to fuck you, then fucking you?

No, hold on. The two articles disagree very much. The one Scott just cited explains that you can't trust a commercial VPN provider.

The Mozilla post says:

> Are VPNs truly private?

> Unfortunately, no. The VPN provider can still log your browsing data. You are essentially putting your trust in your VPN provider. Will your provider hand over info when pressed? Will they log your browser data and sell it at a later date?

Which is basically also saying you can't trust a commercial VPN provider. I suppose it does differ in that it says it's still an option, though.

I trust most VPN services more than I trust my ISP. If what you are trying to do is avoid your ISP collecting your surfing data for advertisers, throttling Netflix traffic, or adding a super-cookie to headers, then a VPN might make sense.

My ISP choices are limited to two companies that are both terrible. A VPN is a nice way of limiting what they can do to you.

You don't get any additional privacy, the only way to really _guarantee_ that you get additional privacy is to use a solution that provides privacy by design rather than by policy.

I'm not looking for a guarantee. Probably getting additional privacy is good enough for me.

> I'm not looking for a guarantee. Probably getting additional privacy is good enough for me.

I think we can both agree that wasting your money on wishful thinking ("maybe provider doesn't log") instead of using free open-source privacy-by-design solutions is a bad idea.

I would certainly disagree with that.

The privacy-by-design solutions have their problems at well (ex: speed). It would be better to use them over VPN IF AND ONLY IF their features would be strictly equal.

As they are not, one simply calculates the expected value of both, taking into account the probability of the VPN actually logging the traffic (which should be low for VPNs with good reputation).

For some use cases, even a VPN that logs traffic would be a good idea. For instance in many countries if you download a torrent they will log your IP and try to identify you. IF you have a VPN, they won't even bother asking the provider the IP because it is just not worth it for something like that. If you were exchanging child porn on the other hand they will ask for it and take time to find you.

Not everybody needs the same guarantee of privacy or has the same risk if the privacy was to fail.

Your statement is the same as saying one should never invest in shares because the return is not known in advance, so you should just buy government bonds which are safe.

How do you not get any additional privacy?

As I mentioned using privacy by design solutions (Tor, i2p, ...)

Now you have to trust two ISPs: Yours and the VPN provider's.

You're thinking of these as Single Points of Failure, but they're not in parallel; they're in series.

Consider the attacker: a service you've visited that has your "outermost visible" IP, and wants to know who you are. From their perspective, it doesn't matter if your ISP is willing to give information freely, because they don't know who your ISP is until they've already gotten the information from your VPN provider. Each layer prevents the layer below it from being attacked, until it is removed.

Yes, a state actor could just ask "every ISP at once" to look at their logs of OpenVPN-protocol traffic and identify the packets that match the ones that arrived at the service. But state actors aren't the usual attacker profile, and require entirely different strategies (e.g. getting human "proxies" to use Internet cafes for you.)

Ignoring traffic analysis, you shouldn't have to trust your own ISP while using a VPN. Ignoring traffic analysis makes sense unless you're a high profile criminal, and it affects all low latency tools, including Tor.

Tor is basically a funnel into the DOJ and has been for quite some time:

https://arstechnica.com/tech-policy/2017/03/doj-drops-case-a...

They run massive PR campaigns with carefully structured press releases designed to convince the kind of people they want to detain that TOR is private and safe for any kind of activity.

Because of this people tend to get swole when you suggest that TOR is not any good for protecting your privacy because lots and lots of people have been arrested, tried and convicted after trying to use it to hide elicit activities.

The US government has made millions of dollars of investment into TOR:

https://www.theguardian.com/technology/2014/jul/29/us-govern...

Pretty much every time the US government is investing in something you can be certain that their intention is not to help you out.

AFAICT, in all current cases it isn't Tor itself that's been broken by the authorities. It's the client end that has been compromised; and in a way that isn't specific to Tor. Had these users been using a VPN without Tor, they could have been compromised in largely similar ways.

Please, find me a counter-example - because I haven't seen one.

Admittedly, one thing that has happened is that the authorities are able to target compromises in the Tor Browser specifically, rather than in a wider range of clients that non-Tor VPN users might use. But they're probably more vulnerable than the Tor Browser is anyway.

It's important to consider here that the average person using TOR is not a network administrator.

And that they'll follow the instructions that come with the TOR browser and assume that it's safe.

So when I say that TOR isn't safe, I mean that it isn't safe as it's presented.

Saying that TOR isn't safe if you know what you're doing is like selling someone a car with no seatbelts and then telling them well if you knew what you were doing you'd install seat belts yourself and then the car would be safe.

> So when I say that TOR isn't safe, I mean that it isn't safe as it's presented.

Sure. But it is no more dangerous to use Tor on its own than it is to use a VPN privacy service on its own. So your claim that the US Government is enticing people into using Tor to entrap them is nothing more than an unsubstantiated conspiracy theory. It would be easier for governments if criminals didn't use Tor.

Chrome is arguably more 'secure' than the ESR Firefox that the Tor Browser is running on. If you are realistically concerned about this type of targeted attack, you should probably be browsing with Chrome isolated inside of Qubes/Whonix.

I meant colloquially. If you're not using your VPN 24/7, you have to trust both at different times.

You are of course correct. :)

My ISP is AT&T. I don't think there's much the VPN provider or their ISP could do to make things worse for me. The worst case scenario is that they are as bad as AT&T and there's a non-zero chance they are better.

That's a shallow analysis.

The worst case scenario is not just that they're as bad as AT&T. The worst case scenario is that they're as bad as AT&T and still provide a false sense of security.

Even if you're diligent, other users with your (ISP, VPN) provider pairing might not be, and they could be harmed as a result.

The comments security nerds make here on HN aren't one-on-one individualized consulting (n.b. that's paid work in my field), they're general advice for the public to refer to.

If you are tunneling all traffic through your ISP, seems to me you aren't trusting them all that much.

I feel like this is dated, because in 2017 this:

> You are on a known-hostile network

is true for every network in the USA. You can be sure they ae all being snooped on by 1. the ISP collecting traffic data for profit and 2: the gov. because they get it all anyways.

I think the most popular use case is torrenting which a VPN will help.

That isn't great privacy wise as it's still privacy by policy. The best way to torrent is to use i2p which - unlike Tor - encourages that activity. (Short tuto: the default Java i2p bundle already comes with I2PSnark, a torrent client. To download a torrent, search through known i2p trackers such as the Postman Tracker: http://tracker2.postman.i2p )

The content owner could still request your information from the VPN provider and the VPN provider might provide it (even if they say they won't). I think the main benefit is that there are so many individuals torrenting copyrighted material that aren't using VPNs that it means you aren't the "low hanging fruit" so you're considered not worth the effort by the content owners.

Yes, but there is a big difference between "this provider might be lying about not storing traffic, and they also might give the data to someone" and "this ISP is 100% storing traffic and routinely gives that data to others."

Why base your privacy on wishful thinking ("provider is probably not lying") instead of using privacy by design solutions? (e.g. i2p for torrenting)

Because privacy by policy is good enough for almost everyone.

> Because privacy by policy is good enough for almost everyone.

Source? And why would it be good enough when it has been shown time and time again that it's ineffective (example: DNT header)?

If you want to torrent, turn one of the low end boxes into a seedbox rather than a VPN server.

Even then, setting up your torrent client to use a proxy is just as simple and effective.

For now, I am running my own VPN on Linode. The only real benefit of this is now my traffic is mixed with non-similar traffic. The hope is that this makes it less valuable to monitor the contents of my traffic. Of course, this just security through obscurity, and nothing more than a half measure.

The internet is not designed for privacy, and privacy does not benefit the majority of commercial stakeholders of the internet. This is probably why most privacy solutions feel like shoving a square peg through a round hole. My personal feeling is that we should combat commercial bulk surveillance through legislative means.

Your last paragraph ignores the existence of many privacy by design solutions such as Tor or i2p. Yeah, they can't protect against a global passive adversary - as any other low latency anonymity system in existence, but that's totally different from saying that there's no way to have privacy on the Internet.

Tor is a solution for specific use cases. It does not address privacy on the internet in a general way. For example, if I use tor to browse facebook, I am logged into facebook and still just as trackable as I would be if I wasnt using tor.

> Tor is a solution for specific use cases. It does not address privacy on the internet in a general way. For example, if I use tor to browse facebook, I am logged into facebook and still just as trackable as I would be if I wasnt using tor.

No, at least now facebook may not know your exact location (especially if you use their onion service: https://www.facebookcorewwwi.onion/ ) and they can't track your activity outside of facebook. Of course, it doesn't solve - nor can any other anonymity system - the fact that you transmitted personally identifiable information with facebook.

Obligatory: https://twitter.com/tqbf/status/700798735190601729

A confusing, content-less, arbitrary recommendation against Linode with no clear justification or reasoning given anywhere in the tweet stack is obligatory? I'm confused. Are there any actual reasons not to use them?

His "recommendation" stems from a DDoS incident, and possibly, a hack.

https://news.ycombinator.com/item?id=10998661

I'm fairly new to whole world of increased internet privacy, so I'm curious of the benefits of using a VPN or Tor. I'm not a political activist or engaging in illegal activity, I just want my personal data being passed around as little as possible (preferably by spending little to no money to do so). Is using Tor worth the effort? What are the benefits? Or do I simply use Chrome and resign to my fate like nearly everybody else?

> Is using Tor worth the effort?

Definitely.

> What are the benefits?

Because of its 3-hop design, a non global passive adversary (GPA) would need to control both your entry node and the exit node to de-anonymize one of your Tor circuits. In addition, Tor circuits generally last for 10min only. Also using the Tor Browser you get stream isolation meaning that you get different Tor circuits for different websites.

You can also setup your own non-exit node and connect to it to ensure that no single point in your Tor circuit controls both the entry node and the exit node.

> a non global passive adversary (GPA) would need to control both your entry node and the exit node to de-anonymize one of your Tor circuits

That's not a benefit, that's a feature. A benefit involves a use-case. What does a person gain from not having their traffic de-anonymized? The described user is someone who doesn't have any particular activities they need to keep secret or risk jailtime. So, for them, what's an example of something that could happen differently in their real life if they used Tor vs. if they didn't?

(This wasn't a rhetorical question; there are such use-cases. I'm just commenting to prod you into zooming out a bit from "privacy is its own end" to thinking more about what regular people care about and how privacy helps them get it.)

For starters, don't use Chrome.

Chrome sends a whole lot of data to Google (and possibly to their data-sharing partners) such as, at the least, what sites you visit and how long you are on each. When combined with Analytics, cookies, profiling and whatever G services you use, and the fact that Chrome is a program (not a site) connecting that all, you have pretty much lost any legitimate hope to privacy before you begin. Use HTTPS everywhere is a no-brainer, as at least the middle steps won't see the data. IMO, using a commercial VPN is just not that difficult and the speed is close to native, so its a lot easier than TOR.

Basically it comes down to this: What you don't want people to know, you don't tell them. So if you don't want personal data floating around everywhere, don't tell them personal data.

Or just be a nice happy good citizen in the normal world. What you do in other worlds should then not be mixed with the normal word.

Most people I know want a VPN to pirate stuff without consequences. So I'd say, Tor would not cut it.

Tor is emphatically not meant for piracy, especially BitTorrent.

As I mentioned in another comment about using VPN for torrents:

> That isn't great privacy wise as it's still privacy by policy. The best way to torrent is to use i2p which - unlike Tor - encourages that activity. (Short tuto: the default Java i2p bundle already comes with I2PSnark, a torrent client. To download a torrent, search through known i2p trackers such as the Postman Tracker: http://tracker2.postman.i2p )

Unless stremio and other pop corn time like can work transparently with i2p, it won't help.

> Unless stremio and other pop corn time like can work transparently with i2p, it won't help.

What? i2p is a self-contained network and not really meant for clearnet browsing.

You need to look up what stremio (https://www.strem.io/) is and understand the value proposal for the casual non tech saavy end user. This is the face of torrenting now. Not magnet links. People don't know what a URL is anymore, don't expect them to understand a classic torrent client or i2P.

Since we're talking about it: what's the value proposition in creating an illegal service for non tech savvy end users?

I'm trying to figure out why they made this. They can't really run ads without ending up like the founder of TPB.

Regardless, it doesn't seem unreasonable to expect people to know what a magnet link is. When all you need to do is download transmission and click on a magnet link, people are fine with that.

You mentioned stremio and I respectably pointed out that it's not going to work over i2p for reasons mentioned above. I don't even see why you're mentioning it when we're talking about privacy.

My whole point is that people use VPN for torrenting so Tor would not help and i2P neither. What are you talking about ? Did you read the first post ?

> My whole point is that people use VPN for torrenting so Tor would not help and i2P neither.

My point was that I2P can help them since it's (a) torrent friendly, (b) has a bundled Torrent client (I2PSnark), (c) there are many eepsite torrent trackers such as: http://tracker2.postman.i2p

But they use stremio which doesn't work with i2p.