Sounds like a lot of work just to listen to random web traffic. Why not break into Facebook or Gmail instead? Then you could spy on anyone you wanted.

Why go to the work of breaking into Facebook or Gmail when you an just exploit this proxy and get Facebook accounts and Gmail accounts and bank web sites and a million other sensitive things too?

Because I would bet that the security professionals working for and contracted by Google and Facebook to be an order of magnitude better at their jobs than those hired by Nokia. I would expect Google and Facebook to be better at closing holes and better at detecting and closing security breaches if and when they do happen.

You know you just agreed with the comment you replied to right?

Oops... I was distracted while reading and didn't read carefully enough.

Because this would only get the information for people who happen to be browsing to those services through specific Nokia phones.

Surely this is a giant target painted on Facebook or Gmail's servers for any blackhat out there who wants to intercept a whole lot of traffic.

Facebook and Gmail don't handle your connection to your bank.

Apart from that, we already know the Chinese have hacked GMail accounts on several occasions, so Google is already a definite target and the Incidents that have been reported may be just the tip of the iceberg so far as we know.

We also know that there have been cases of corporate and government espionage in the recent past. This is happening, right now, and we may never know how extensive it is. It's entirely possible that Nokia could become a target of either technical or even employee infiltration. If so, covert access to a service like this would be a fantastic coup, especially since Nokia phones are still popular in many parts of the world with unscrupulous and technicaly sophisticated governments.

Despite the irony, I'm not arguing it should be less of a concern.

Recent mishandling of user accounts by Sony Entertainment and LinkedIn gave us a glimpse of how much damage an intrusion can cause, even though (?) their management thought their intended purpose couldn't lead to any serious liability (speaking through my hat here).

Of course it is. But that's beside the point. The fewer servers that I interact with that have giant targets and potential security holes the better.

Facebook and Gmail will not have bank details or credit card numbers sent via them.

These servers probably do, as I'm sure some people use them to do online banking and shopping online.

You'd be surprised what sensitive details people send via Facebook messages, email and other insecure communication tools.

No more so than any other proxy server.

Can a normal proxy server decrypt your HTTPS even it if wants to? I'm by no means an expect on cryptography, but I assume they're only able to do this because they can include their keys on Nokia devices and tell them to trust it.

When I access a proxy on my computer, it can't do that. If an arbitrary proxy could MITM a secure connection, HTTPS would be useless.

No, an ordinary proxy cannot do that. HTTPS is supposed to be end-to-end. I.e. the communications can only be decrypted by the two endpoints. One of the main reasons for signed SSL certificates is to prevent a middleman from masquerading as the endpoint and convincing your browser to negotiate encryption with the middleman rather than the real endpoint.

So I presume the Nokia browser is complicit in this scheme.

In the end, you have to trust the browser.

google chrome sends all my pages to translation and what-not. I have to completely trust it. that's why i never use the compiled version but the chromium one only (hence not having access to any addon via the addon site, have to do some manual work there)

Even besides the browser, how many computers don't I see the skype button next to phone numbers? would you trust skype is behaving and not sending your data to their servers? did you remember to disable this add-on for ssl pages?

You're conflating different issues.

A browser based on an open-source codebase with many people auditing its source-code and network traffic (there was much paranoia about Chrome when it was released). And even in that case, you have the choice to use a completely open-source browser that has a different stance on user privacy (Firefox).

An NPAPI plugin that can be easily disabled.

A third-party BOX MITM all your secure connections without your knowledge.

I don't think the above are comparable in magnitude.

You're right--the third is the worst by far. And this proves that you can't trust a pre-installed, commercial browser until it has been thoroughly audited by independent researchers.

They don't decrypt HTTPS. The NokiaBrowser (former OviBrowser) is a "proxy browser", a different application that talks to Nokia servers and these servers than establish a regular HTTPS connection. The same as Opera Mini. There is one more layer of abstraction.

Do you trust your phone operating system? How do you know it is not capturing your data when you access a bank website? Same issue here.

The one difference is that the Nokia servers might be a bigger target for hacking, instead of hacking individual phones.

If the proxy also gives you certs to trust it can. Many corporate proxies do this (silently, as they control the PCs) and I've seen more than one academic campus where to access the wifi you have to accept a certificate.

On iOS this comes up after connecting to wifi as a reassuring, secure-seeming, official-looking page that pops up inviting you to accept the new certificate. Do so and boom, you're vulnerable to a MITM.

Unless you control the CA certs installed in the browser, which nokia does, you can't generally decrypt https connections, which is the issue here. It's a man-in-the-middle attack but it's being used "for your own good".