Hacker News Reader

Hacker News

Comment by andix | Hacker News Reader

Parent Full thread

andix

VPN providers might be much more trustworthy than other ISPs.

View on HN

Or they might be much less? What is your argument?

They have a higher amount of customers who think they need protection. They are an obvious target for surveillance.

And as TFA says:

> VPN providers that lied about "no logs" never faced any meaningful consequences. https://www.theregister.com/2011/09/26/hidemyass_lulzsec_con...

I have no idea what "TFA" refers to.

My argument is: Some/many providers are known to abuse user data, are required by law to store them. Are actively and legally selling the data.

Most VPN providers don't do that. Some might, but a lower probability is still better than knowing that the ISP is legally required to store your data and give them to the authorities.

(this really depends on the country, in some countries ISPs are required not to store connection data, in others they are)

TFA = The Featured Article… or something.

If I ran a spy agency or specialised in blackmail the first thing I'd do would be to run a VPN service. We have no way of auditing their trustworthiness.

If VPN providers would do that, it would soon be public knowledge.

For some ISPs the abuse of user data is a known fact. So VPN is just the safer bet. But don't get me wrong, I'm not claiming this is impossible, it's just less likely.

CryptoAG was owned by spooks for decades.

What makes you think VPN is different.

https://www.washingtonpost.com/graphics/2020/world/national-...

It might be true. But I know that the shady cop in my neighborhood could easily put in a request to my ISP to get some data about me. But he could never get the same data from the NSA. I'm not a terrorist, so I don't feel threatened by the NSA. The thing I need to protect myself against is corrupt law enforcement.

So you're saying you have nothing to hide? Gotcha.

Please tell me how to do it right.

How long did Israel’s supply chain attack on Hezbollah’s pager source stay clandestine?

How long did the USA keep it hidden that they owned the Swiss company that manufactured encryption devices for other countries?

They are extremely adept at finding ways to use the secret knowledge without blowing their source. A good example of this is the scene in _The Imitation Game_ after they break the Enigma code for the first time, but Alan Turing explains that they can’t notify the allied ship convoy without leaking that fact.

As long as you are not a terrorist, all those surveillance programs will never be used against you. It would put those programs to risk, by exposing them.

If you need to protect yourself against the most sophisticated surveillance programs, a lot of opsec will be required.

VPNs can protect you against some creep in a public WiFi spying on your DNS request. It can protect against some low-tech state run mass surveillance.

Good thing the US doesn't have an administration that will happily criminalize any opposition

If this is your threat model, then VPNs are definitively not enough. Still a European VPN provider might be better than nothing.

> As long as you are not a terrorist, all those surveillance programs will never be used against you

Can you say "as long as you have nothing to hide"?

We’ve known since WW2 and maybe further that governments have used surveillance against innocents and people they didn’t like.

So you're suggesting to use Tor for everything? Or what's your solution?

All those surveillance programs are a threat to everyone who wants to keep the technology he and his company develops.

There is nothing preventing programs of that sort being used for industrial espionage.

This may be a useful notion and term to start using when thinking about these things: https://en.wikipedia.org/wiki/Threat_model (not being smirky; I should use apply it more often myself, e.g.)

With this in mind, I'll link my prev comment: https://news.ycombinator.com/item?id=43095936

> Israel’s supply chain attack on Hezbollah’s pager source

You know, James Mickens, a CS researcher at Microsoft had a very related article (PDF), "This world of ours": https://www.usenix.org/system/files/1401_08-12_mickens.pdf

See Figure 1. He calls it the "Mossad / Not-Mossad threat model" :) (but, joking aside, it's good to identify oneself on a scale, or expectations for some system scale-wise and property-wise (incl. which security properties, esp. privacy vs. anonymity!) before looking closer at it.

Anybody expecting to get enough privacy to commit crimes using a VPN is definitely naive. I think most people doing important privacy stuff like being whistle blowers in authoritarian regimes knows that well enough. As part of a multi-pronged strategy to reduce the surface area exposed to creepy data brokers, though? Works fine for me.

Exactly the point. It's a security measure for regular people, who don't commit crimes and aren't a high value target for government agencies.

With a VPN your employee can't track what websites you're visiting over the WiFi in the break room (looking for a new job, visiting the union website for legal advice).

The hotel can't track what you're doing on their WiFi.

While traveling to not really democratic countries the government can't spy on you (in many muslim countries the typical US/European browsing behavior is probably illegal).

Right. A lot of folks sacrifice the good because it’s not unattainably perfect in tech, especially with security. I think the pressure to improve, on a whole, is a positive thing, but it often makes terrible practical advice.

> regular people, who don't commit crimes and aren't a high value target for government agencies.

> While traveling to not really democratic countries the government can't spy on you (in many muslim countries the typical US/European browsing behavior is probably illegal).

Aren’t these two threat models already completely different?

IMHO if an authoritarian government considers what you’re doing to be a crime, you should not be doing it over a VPN you saw advertised on a billboard

I think they’re mostly referring to bypassing countrywide web filters.

If you use a VPN provider that only has one set of servers provided by one hosting provider then sure. If you’re connecting to any number of servers then that’s probably less of a problem. I’m much more concerned with Starbucks collecting data on their free WiFi than one of a number of hosting providers or whoever I sometimes exit through. The services you log into selling that data to the same brokers as everyone else is a much bigger vulnerability IMO.

Or less, you really have no way to know. You can't verify anything they say.

Some VPN providers publish third party reports about their security. My VPN provider does that very extensively. My ISP doesn't do that at all.

It's about who you trust more. Your decision to make.

The problem with those reports is they still mostly come down to the auditor taking the provider's word for it.

Maybe the staging environment the auditor looked at doesn't log anything, but what about production? Maybe it doesn't log anything today, but who says it didn't yesterday, and that they won't turn logging (back) on tomorrow?

At least where I live, there is strict legislation governing what ISPs can or must collect, and what they do with that information. Offences under that telecommunications interception legislation come with prison sentences, and so there is a strong incentive for individual employees to comply.

For some people, moving that exposure to a VPN provider mitigates their risk. For others, it increases it.

Reply on news.ycombinator.com