Slashdot Mirror
Nokia Admits Decrypting User Data Claiming It Isn't Looking
judgecorp writes "Nokia has admitted that it routinely decrypts user's HTTPS traffic, but says it is only doing it so it can compress it to improve speed. That doesn't convince security researcher Gaurang Pandya, who accuses the company of spying on customers."
From the article, Nokia says: "'Importantly, the proxy servers do not store the content of web pages visited by our users or any information they enter into them. When temporary decryption of HTTPS connections is required on our proxy servers, to transform and deliver users' content, it is done in a secure manner. ... Nokia has implemented appropriate organisational and technical measures to prevent access to private information. Claims that we would access complete unencrypted information are inaccurate.'"
security researcher Gaurang Pandya
What are this guy's credentials apart from being a guy with a blog?
Amazon Silk browser does the same, Opera mini does the same, what's with this jumping on the Nokia hate bandwagon? Perhaps they should stop proxying HTTPS traffic, but remember in third world countries data comes at a HUGE premium, so these services are a god send, especially with a lot of sites moving to HTTPS by default. I would hope that Opera/Amazon/Nokia are atleast as credible as your ISP though it's an additional point of failure.
This space for rent.
Yes, we're opening your mail, but we're not LOOKING at it. We're just making sure you aren't wasting paper and ink.
We don't access your personal information with our closed source NSA backdoors, we just plug this strange Narus device into our routers.
The reason Nokia is able to do this is that they control the browser. According to the article browsers on Nokia phones are delivered with a certificate, that allows Nokia to perform this MITM attack. They call it a feature and provide a plausible explanation of what benefit it has for the users. However enabling such a risky feature without user consent is a really bad move and means users should no longer trust Nokia products as much as they have done in the past.
Do you care about the security of your wireless mouse?
There must be serious flaws in HTTPS if they can decrypt the traffic for hosts that they don't control the certs for.
They control the browser. According to the article, the necessary certificate is installed on phones as Nokia ships them.
Do you care about the security of your wireless mouse?
On their own phones, they just install a browser and their own trusted wildcard cert.
Then anything you browse to, the browser trusts and encrypts but just to the "wrong" destination.
On any decent machine, or decent browser under your own control, you wouldn't let it happen. And if you did, SSL would be similarly "broken".
SSL is a trust mechanism only. If your phone trusts Nokia, the padlock icon means nothing beyond that you're talking to Nokia. If your phone DIDN'T trust Nokia, it wouldn't be an issue and they would have to pass your traffic through unchanged (and still encrypted!) to the destination servers or risk SSL warnings on your browser.
This is why you don't ignore browser certificate warnings, and why you NEVER install a certificate on your computer (or allow software to). I've seen software that installs a trust certificate for the vendor when installed (as administrator), that would be show up and be allowed in the IE certificate store too (so browsing to any site with a cert signed by that cert would let you think you were talking to Google, etc.)
See also Google's TURKTRUST issue lately - if you trusted TURKTRUST, you thought you were talking to Google and weren't. If you didn't, you would just have got an error and still been secure.
There must be serious flaws in HTTPS if they can decrypt the traffic for hosts that they don't control the certs for.
I guess if Nokia controls both the proxy server and the mobile device then their implementation of HTTPS can be designed so that the mobile device trusts the fake cert on the proxy server.
There must be serious flaws in HTTPS if they can decrypt the traffic for hosts that they don't control the certs for.
They control the browser. According to the article, the necessary certificate is installed on phones as Nokia ships them.
This is exactly what i was thinking/fearing. This is some scary shit, basically you ought to treat HTTPS on your Nokia device like HTTP, unless you really really trust that Nokia knows what they are doing and how to keep a secret. The striking thing is that users obviously have no idea they are handshaking with Nokia instead of their bank, doctor, etc. Are there at least alternate browsers available?
Isn't that the whole point of HTTPS, to ensure that a man-in-the-middle attack (in this case, a probably benign proxy) is impossible?
It is only impossible without the collusion of a trusted certificate authority. When was the last time you reviewed the list on your browser? Oh, and did YOU do anything to determine if any of those organizations were trustworthy.
If you get a mobile device from your mobile provider, there is a pretty good chance that they stuck their own root CA in there somewhere. Maybe they just use it for SSL connections to their own websites/email/etc. But, trusted is trusted in the world of SSL which means they could just MITM every connection you make.
Ditto for any PC you use at work. Chances are your employer has a trusted CA somewhere in there, which means they can MITM any SSL connection you make to any service on the web.
If they didn't actually modify your browser you can probably spot this by pulling up the certificate info for your connection and noting who issued it.
This is why I believe SSL offers a false sense of security. Moving to certificates distributed over DNSSEC would cut out the middlemen, and it would improve security. Only the domain registrar for google.com could tamper with their certificates, for example. That still isn't perfect, but it is better than any CA anywhere on the globe.
Mine does (Australian government department). Interestingly they specifically exclude the local banks.
Dear god. Is this what corporations do instead of serious engineering work to debloat the network stacks, drivers and hardware or start implementing things like TCP Fast Open? :-|
Another example where fixing bufferbloat needs a strong front because people will start doing the wrong things when trying to fix something.
Just as BitTorrent-induced latency was made the culprit of slow networks and caused people to think it's good to go away from Net Neutrality and charge premium for a premium experience. Nonsense!
Wrong profile linked. Correct profile. Stupid misclick. Ugh. In other news, his background is not a software developer, but a network admin with some cisco experience. Like many in that area of IT, there is some exposure to security. I wouldn't call him an expert in MIM attacks, but he's not a layperson either.
#fuckbeta #iamslashdot #dicemustdie
The same thing can be (and is) accomplished in normal desktop OSs by adding a CA certificate to the certificate store. It's commonly used in businesses that have an HTTPS proxy as well as an HTTP proxy so they can filter/monitor HTTPS access as well. IIRC there was an Ask Slashdot question about it as well. In any case, no modification of the implementation is needed.
...my ass
Right up until the government shows up and demands that they send all the traffic to them first, and forbids them from notifying their customers.
If you're using BES, it's all encrypted - it goes through RIM's servers, but RIM can't read it.
Hence the big kerfuffle about governments insisting on access to BES data, and RIM's refusal to give it -- they literally can't.
Consumer email/BIS access is a different story. RIM does have access to that, and presumably government as well (similar to what any other provider gives).
How is easy, as other have said. How legally? That is another matter. As I read it, they are committing a DMCA violation by breaking a security measure. Should be able to go after them for anticircumvention tools, and force them to remove the cirt.
This is some scary shit, basically you ought to treat HTTPS on your Nokia device like HTTP, unless you really really trust that Nokia knows what they are doing and how to keep a secret.
Any web page retrieved through HTTPS is parsed into an unencrypted DOM within the web browser. You have to trust that the browser publisher knows what it is doing and how to keep a secret.
HTTPS is only as secure as the implementation. The implementation in their browser deliberately implements it poorly, and accepts Nokia's server saying "yes, I verified the certificate on the remote server" as being valid verification of the cert.
Uh, my ISP can record all the SSL connections they want, because they can't decrypt what I'm sending.
So are Nokia spending their Microsoft billion on astroturfing Slashdot, or does it just look like they are?
Wasn't it Benjamin Franklin who said "They who can give up essential security to obtain a little speed increase, deserve neither security nor speed"?
Your trust is extended because of the expectations involved. The user/owner of the device is not informed that, unlike his PC or other smart phone devices, Nokia is handling encyption differently. As https is used primarily for the purpose of securing data traffic between the user and their banks or their other services which need security, the expectation has always been that it would not involve the maker of the device which is being used.
I "trust" my car maker to build a good car. I do not "trust" them not to install cameras in it without my knowledge and then tell me later "there are cameras, but we are not looking at the video feed."
It may be illegal in the US as well
Just like warrantless wiretapping...oh wait!
"A person is smart. People are dumb, panicky dangerous animals and you know it." - K
I don't trust Microsoft in the slightest, but I can use their stuff on my PC because I have the ability to audit and control what comes in and out of my computer. If they try something, either I can discover it myself, or one of a hundred security researchers will be able to find it. Also, the application software encrypting my data is installed by me and under my control and ability to inspect.
The idea with HTTPS is that you know that you *cannot* trust the intervening internet/cellular carrier infrastructure to not be monitored, so you set up an encrypted discussion that can pass through that untrusted domain without being read. Nokia subverting this process for any reason, any reason, renders it pointless because Nokia is now a third party that can read your data, even if they double pinky swear that they won't be evil. I don't want their assurances, I don't want them to even be able to do it, period.
I imagine that most people did not realize that Nokia had subverted the certificates and they think that they are having a more or less safe conversation with their destination... as they would be if Nokia didn't replace the certs.
Not really, it's relatively trivial to establish a man in the middle attack if you completely control the communication channel. A requests a secure channel to B from C. Instead C establishes a secure channel with A *claiming* that it's B, while also establishing a secure channel to B claiming that it's A. Theoretically any node your connection passes through could do this, but given the fluidity of internet routing algorithms only the ISPs at either end are likely to be able to actually pull it off. Or any routers between them and the actual computers that are doing the talking of course.
That's why they tell you never to do internet banking, shopping, etc. at an internet cafe or other open hotspot - a fully controlled malicious data channel can do whatever it wants, and how are you going to detect it? All the validation has to go through them.
In the case where you have vendor-controlled browsers or proxy servers it's even easier, but basically those are just additional nodes your data is guaranteed to pass through.
--- Most topics have many sides worth arguing, allow me to take one opposite you.
If you don't trust Nokia to not snoop on your data then why are you carrying around a device made by Nokia that contains a camera and a microphone and a cellular connection to the internet (and probably a gps though I don't know the details of Nokia's phones)?
They don't just tell you - they advertise it. It's one of the phones biggest selling features.
The issue in countries where the phone is sold is network traffic. It's costly. VERY costly. This browser does what opera mini did for about a decade - it works through nokia's special proxy that fetches the page for you, renders it in unique way that saves a lot of traffic and then sends it to your phone's browser.
The user makes what he believes to be an encrypted connection. Nokia interposes their server into this connection without the user's knowledge and decrypts their data (both ways), and then claims this is perfectly OK, since they're doing it to optimize bandwidth or such. whether they make use of the information or not, they are intercepting and decrypting a connection the user believes to be private.
This seems awfully like wiretapping and unauthorized interception of data communications. If it isn't illegal to decrypt an encrypted transaction if you are not the intended recipient, perhaps it should be. I'd wager it *is* illegal under EU data protection laws, but IANAL. It's probably OK in the US, due to some obscure law permitting just this activity, passed at the request of some large corporation.
Yes, this IS wiretapping. I don't care if they've got a tiny tiny line item in their terms of service that say they're doing this, NO ONE expects their https encrypted session with their bank to be in the clear on Nokia's servers.
I'd really really like to see the RCMP charge Nokia Canada's CIO just on principle. Just because big companies have lawyers and huge t.o.s. don't mean they should be treated any differently than joe blow secretly inserting software on his aunt's computer to listen in to her voip conversations.
They advertise the feature without advertising the implications.
Of course, that's called "marketing". Push up the upsides, burry the downsides.
Write boring code, not shiny code!