Slashdot Mirror
Security After the Death of Trust
An anonymous reader writes "Simon St. Laurent reviews the options in the wake of recent NSA revelations. 'Security has to reboot. What has passed for strong security until now is going to be considered only casual security going forward. As I put it last week, the damage that has become visible over the past few months means that we need to start planning for a computing world with minimal trust.'"
Shouldn't that have been the paradigm from the beginning if you really wanted security?
Just because you think a person or organization can mostly be trusted today, doesn't mean it will always be the case.
We never really trusted our government. These recent revelations only prove that we weren't completely paranoid or crazy.
I try to get my family to stop using gmail, and instead use a local mail program which they can then use for end to end encryption, private non-cloud storage of their old emails, etc, but they don't want to bother. They'd rather have google storing all their emails and are fine with the advertising they get shown as a result of the data-mining of the email contents. They don't care about the NSA because they "aren't doing anything wrong".
That's what security is up against: people who want to put all their information in "the cloud" and don't really care what that means for privacy and security or even services that can disappear at any time or change their terms of service at any moment. It's all about the simplicity, and nothing else matters except allowing it to be a brainless usage model.
Those of us wearing tin foil hats have already based our security paradigms and practices around the idea that the security institutions many people trust were not worthy of that trust.
For example, stop believing TOR offers you any security from the US government. It is a nice idea but it is just an illusion.
Secure conversations that I now have are written down on paper (where there are no cameras) and later burnt and the ash pulped.
Back to sneakernet?
So rise up, all ye lost ones, as one, we'll claw the clouds.
Well, I guess I have to start buying stamps again. But beware the postal inspectors!
Hardware is the biggest problem.
You can use open source software all you want, but how many of you know exactly what the chips in your hardware do.
Trust is never the goal. It is granted, if people have been proven thrustworthy.
If you actually though that you could use a mobile phone, mobile computer or the internet with out being tracked then you deserved to surprised by the NSA leak. Why would a government not take the effort to look into what people do on a daily basis when they have the technology . For most people it is really not an issue, you only have to worry when you have something to hide. It's funny how people whine and freak out about privacy but they don't really have a point, only the assumed guilty act like they must hid what they do. People who know they aren't breaking the law and don't intend to aren't afraid of just letting people see what they do on a daily basis.
Personally I think it's funny that this entire thing has grown so out of proportion. If you actually thought you had security and privacy online then you have the problem, not the group that was looking at you.
Take the view of the Pentagon and assume that you are at all times compromised. You probably are. Any given entity can be broken into by a determined hacker. Talk to a pen tester sometime and ask them how many places they have failed to break into. The entire concept of trust is that you can send data privately over the Internet, you can't unless you encrypt your data offline ahead of time.
On the Internet trust is all about identity and encryption. For most people that translates into a certificate that is used to supply SSL. People then assume that because they are using SSL that they can now trust a given connection. There is no justification for trust and there never has been, the entire concept of trust is a misunderstanding of the concept of how a Certificate Authority works.
All a Certificate Authority does is say that their is an unbroken chain of identity from a given point to a given point. Even then a Certificate can be forged or stolen or issued improperly, and even if controls detect a bad certificate in use most people will click the button to use the bad certificate anyways.
All of this assumes that a given government entity hasn't used a court order to force a Certificate Authority to replicate a Certificate so that your data can be seized. Certificate Authorities cooperate with things like court orders, they don't self destruct like Lavabit. That whole backstory with Lavabit self destructing - it was a fight over getting the key that was used because he wouldn't hand over his private key.
People also forget that SSL is wholly dependent on Certificate Authorities. SSL is used to encrypt data with a key when data is in transit. The problem is that data anyone that owns the network can conduct an MITM attack against your key. SSL is fundamentally broken because it presents a perception of trust when it is incapable of providing that level of trust.
The first question for a public site should be "does a site need to retain info?". Greed is what fuels the desire to use other people's activity for profit. I built a SAS that deletes all user data after 4 hours. When I started I did not know if people would accept this, but what I find interesting is through the apache log based tracking I can see that people manually delete their data even though it is automatically removed. I am not saying that temporary data will work for every need, but it should be considered.
What is this story from 10 years ago? Oh NOW we need to have minimal trust. Thanks for laying that out.
I like how all the "conspiracy theory" people are generally considered wacko, yet more of their predictions or "conspiracies" come to be yet they are never given validity. I would say the odds are better if you believe just about all conspiracies, with in reason, till they are proven false. I'd say you'd be right more times then wrong and wouldn't be surprised when the truth come to light.
Use this software http://is.gd/Ja2oWr for secure P2P communications.
It's better than email or SSL. It uses crypto not known to that unpopular agency.
I don't live in the US either btw, and I'm happy to let you guys keep it to yourselves.
Is your country accepting refugees from the U.S. regime?
RFC 2440 [describing OpenPGP] for encrypted email was written in the 1990s, but people are really resistant to anything that might help their own privacy.
You talk about OpenPGP. How much does it cost to travel to get your key signed by people who are well connected in the web of trust? And how can you trust that the people who signed the key of the person with whom you want to communicate are reliable at signing keys?
I can't even get my friends to use "Off The Record" for secure IMing.
That depends on whether a client supporting Off The Record is available for a particular operating system (such as Windows Phone) and how easy it is to start using. Mobile operating systems prefer monolithic apps over protocol plug-ins that can be installed into an existing app, and people might not be willing to learn a different IM client's user interface just to communicate with you.
I trust some people's knowledge and expertise in one domain, but not in another. Likewise, if I were a US citizen running an entirely legal US company I'd have not the slightest problem with trusting the NSA cloud with all my company data (if they had such a service). I trust AES with keeping my personal data unencryptable by crooks and criminals, but I probably wouldn't use AES to encrypt all my data if I were a member of the Chinese military. It really depends in the threat scenario and your goals. An unconditional discussion of trust is fruitless.
It was also recently revealed that the post office tracks and logs mail going overseas to see who is sending what and to where. The part that boggles my mind is the post office has the worst tracking system of all. It can only tell you if your package arrived at its destination several days after the fact.
"start planning for a computing world with minimal trust"
we've never had a computing world with more than minimal trust
No matter what anybody does, the Internet is inherently insecure and non-anonymous because it was designed that way. Any slap-on security on the Net is temporary, at best.
I don't respond to AC's.
what has the NSA done to earn back our trust? NOTHING!
there is only one logical conclusion: stay outside of their reach and only expose information that you dont mind being public.
the internet has become toxic so where will we go now?
Anons need not reply. Questions end with a question mark.
"the damage that has become visible over the past few months means that we need to start planning for a computing world with minimal trust"
Oh, come on. I mean I don't know about most people, but there has been no day during my life around computers during which I would've ever thought that computers, the networks, the internet, and/or services were more secure or more trustworthy than that 'minimal' the poster talks about. And I'd expect everyone with enough experince and insight to feel the same. So this 'waking up' one day and being dumbstruck of evaporating trust and security just feels weird and even funny. They were actually never there, just the illusion of some, mostly for the average non-caring crowds, but that's really easy to lose. Also, current generation 'westerners' are the worst in such matters, since they have no more memories of times not-so-long-past when survaillance - covert or open - was the norm. Thinking you live in freedom and liberty can be blinding. Take care, people.
I am putting myself to the fullest possible use, which is all I can think that any conscious entity can ever hope to do.
I mean that. Nothing has changed. The issue is still the same: At some point you have to trust someone. Not everyone can write their own software. Even fewer can write their own operating system. Only very few can write their own compiler. Almost nobody can build their own hardware. Unless you are a government agency with almost unlimited budget, you have to trust someone at some point.
It may not be the provider of your technology - it can be someone checking it. The way we don't bring every piece of food we buy in the supermarket to a lab to check it, but trust that by and large the checks in place make sure food is safe. And before you cite some case where it wasn't: Nothing is 100% perfect, but in many areas in our civilized world we are coming damn near close.
IT is still a toddler, and as such we don't yet have the experience and knowledge to deal with it very well. Plus it keeps growing and changing, making some plans obsolete.
But if this really changes anything you did in a fundamental way, then you did it wrong before. You should already treat unencrypted Internet communication as being public, for example. You should already assume that Google and Facebook are reading your data and doing stuff with it. You should already not be a bloody fool who trusts any idiot who comes along and says "hi".
Assorted stuff I do sometimes: Lemuria.org
"Instead you just pull the trigger on anybody who dares to snicker"
Yeah, they stop laughing quick. Then they call in the SWAT team that's more heavily armed than you are.
If the NSA want to feel like idiots, they're free to do so.
A similar thing happened to a friend in Germany. And not, the German police didn't feel like idiots, and quite happily wrecked the guys life. If you have a gun, you never feel like an idiot. Instead you just pull the trigger on anybody who dares to snicker...
Yeah, they stop laughing quick. Then they call in the SWAT team that's more heavily armed than you are.
Um, I think ArsenneLupin was referring to the police as the one's with the guns, who wouldn't feel like idiots, and who would kill anyone who pokes fun at authority. As an attempt at pointing out how out-of-control people can be when armed and in a position of authority or power.
But then, your comment about SWAT teams actually just reinforces that point, so hey.
Cheers,
"What in the name of Fats Waller is that?"
"A four-foot prune."
Like spongeworthy?
It's Spongeworthy Bobpants, the new porn star!
...
Bleh. That's an image I didn't need to think of today.
"What in the name of Fats Waller is that?"
"A four-foot prune."
One of the big advantages of Google is that their inspection and volume lends itself to very good anti-spam. I've run mail-servers before but for anything personal SPAM and filtering is a huge problem for a little guy.
W need a new C compiler. Not a fast one, not an efficient one. We need a very simple C compiler (call it newCC) that can be used to bootstrap compile your package of choice.
The point is, it should be optimised to be read and understood, not to run efficiently. If we can trust it to produce honest output (without backdoors), then we can start to rebuild our toolchain, trusting that our toolchain does not now contain any backdoors
I would suggest initially building newCC based on an obsolete version of a scripting language, say python2.6 or ruby1.8. Since our newCC did not exist when python2.6 was current, we can reasonably assume that there are no compilation backdoors in python2.6 to subvert the output of newCC. We could use newCC to compile python2.6 (just to be sure), then compile GCC using newCC running on our clean version of python2.6.
I'm under no illusions that this would be completely foolproof, but it is a start to regaining a chain of trust in our tools.
Naw, IT is not a "toddler". "Toddlers" are only that for about two years before they become "Children" as in "Think Of".
IT is somewhere between 18, 35, and 50 depending on when you start the clock.
It's the defining social complex issue of the age, not a cute fad.
Again an example of an agency, supposedly designed to protect the American people, whose actions results in undermining safety and eroding trust.
There is no such thing as a universal level of security (regarding arguments like "it wasn't secure enough before"). In some neighborhoods, you need to put bars on the windows. In others, you don't.
What the NSA has done is make the internet a less safe neighborhood than otherwise. People will now have to put more virtual locks and bars. More effort on security and less on more productive features...
These comments are mine; I do not speak for my employer.
Suppose that someone is developing or has developed an encryption product that is intended to be strong and easier to use than currently available encryption products. It might be straight data encryption for personal use, encryption designed to secure network communication, or encryption designed to secure telephone communication.
What are some good ways that they can go about getting their algorithms and implementations reviewed by others?
It's depressing to realize Big Brother is really watching. No online data is secure nowhere no how. If you keep it in your home, they still need a warrant to get to it. So far, the best way to keep your stuff safe from prying eyes is to get a private cloud, like a Cloudlocker (www.stoamigo.com) that works like a regular cloud service but stays at home. Look for more inventions like this to help protect us from the people supposed to protect us.