While I agree that its horrible for agencies to be doing this, I must also say - if it is breakable, then it is not good enough. We shouldn't have to worry about whether someone can break encryption or not. We shouldn't have to guard the guards. The factors of morality and ethics and trust needs to be removed from the equation entirely, because there will always be parties that are not moral, ethical, or trustworthy. This is happening at a good time in history. There are people who are doing what they can to make sure it doesn't happen again. We've only had the internet for 30 years. Lets patch it now.
The title is somewhat unclear, but the issue TBL has is not the cracking itself — as he notes this is a risk and race the internet normally faces — but the deliberate weakening and introduction of flaws into cryptosystems for the purpose of later snooping:
> it's naive to imagine that if you introduce a weakness into a system, you will be the only one to use it. A lot of the IT industry feels that's a betrayal. [...] The two governments have elevated the fight against organised hacker gangs and militarised cyber-attacks from states such as China to the rank of a top national security priority. Yet at the same time their spying branches have actively aided cybercriminals by weakening encryption.
and the effect this has on users through lowered confidence in the privacy of exchanged information.
Then the agencies are endangering their own countries. It is simple math - the world produces much more geniuses that NSA employs (or GRU or anyone). If you deliberately introduce weakness someone will find it. And the entity may not be friendly.
The problem comes because the win of the cold war somewhat skewed the western view of their own capabilities.
When NSA had adversaries that they deemed worthy - they took the effort to strengthen everyone's communications. Right now while looking at the others with disdain and (wrongly) refocused the efforts on terrorism - they prefer weak networks.
Of course there will always be entities who try to break the system - either through technical, political, or social means. Fortunately the existing agencies who had enough power to do this have lost their credibility, so I doubt they will be steering people back anytime soon.
I am not sure if what I said above is even possible, but I hope it is. Trying to remove the side effects of humanity from something created by humans is not an easy task.
There is no free lunch. Securing your communications against every possible attack has a cost. If the cost is so high that even the intended recipient cannot pay it, that may be as bad or worse than transmitting in the clear.
It will always be an arms race between defenders and attackers, because the cost of everything changes over time. You will always have to worry about some genius getting struck by a satori that makes your entire encryption scheme worthless.
The problem is akin to someone in the shipbuilding business worrying about the discovery of a beam that can reduce the strength of a 1-cm thick steel plate to 1-cm thick toilet paper--and the effect is not attenuated by normal matter. The inventor could build a facility in the hinterlands that could instantly and tracelessly sink any steel ship, anywhere in the world.
That is what encryption designers are up against. TBL is upset because intel agencies have been quietly quashing non-vulnerable schemes and/or actively promoting vulnerable ones, paying no heed to the possibility that their enemies might also have some geniuses, or that their own geniuses' ideas might replicate and leak outside of their info-fortress.
"Internet security is hard," he says with emphasis. "All systems have undiscovered holes in them, and it's only a question of how fast the bad guys can discover the holes compared with how fast the good guys can patch them up."
The internet should be thought of as something that transcends the notion of governments and nations; it's a tool for all of humanity regardless of who they are and where they live. The NSA should be considered just another "bad guy" in this scenario - a well-funded arm of a group of people who don't necessarily have the good of humanity as a whole at heart. Literally anyone who wants to manipulate the way the internet works away from an open tool for communication and sharing, to subvert it and break it for their personal gain, or that of their corporation, organisation, government or nation, is the bad guy.
I do find this an interesting notion. Because the internet transcends national boundaries and government jurisdiction, it encourages people to think of themselves as part of a global humanity, without loyalty to a specific government or group, and without loyalty to those who try to invade privacy for corporate or nationalist reasons. It's easy to forget that the concept of nation states and nationalism is a relatively new one and not necessarily worth of respect. Perhaps the internet and sharing and protecting our own data will help to obsolete nations and encourage people to be loyal to ideas, not brands or polities.
Nationalism is mostly pretty evil IMHO, but nation states are actually providing a crucial role as a check on the power of corporations, the boards of which would otherwise be defacto global lords. The tension of diplomacy between nation states is critical to keeping everyone honest. My biggest fear for the future of the planet is a consolidated world government with real power (as opposed to the United Nations) because I think serious corruption would be unavoidable over time.
And among the hard things is the fact that you can't have it both ways. In TBL's case this means he has to realize that DRM is Big Brother Inside. Standardizing DRM means building a standard framework for back doors. You can't be an enabler for the big entertainment companies without being an enabler for snooping.
Many security experts have not faced the fact that they cannot protect their users from state actors and still be deputized by the same state in crime-fighting. The only way to protect the user is to put the user's data out of reach of both spies and police. The security priesthood is easily co-opted and turned into witting and unwitting tools.
Any country that tries to create what he calls a "walled garden" of the internet would find the value of its GDP drop through the floor. Trade would be disrupted, cross-cultural exchange wither
While I agree with the sentiment of the article, I can't help but look to China's version of a "walled garden of the internet" and how that has affected China's GDP. It seems that China continues to do quite well in spite of the aforementioned "walled garden" approach to the internet.
Is the rest of the world doomed to a similar situation?
The whole cyber arms-race makes no sense. It's like mutually assured destruction when you have the capability to remotely dismantle each others' nukes by publishing your designs.
This feels pretty rich coming from the same guy who kowtowed to the copyright industry in order to bring us baked-in DRM on the web.
In practice, the singularly boneheaded decision to allow DRM into the official HTML5 spec will have a far greater impact on the average user's web experience than NSA backdoors will.
Go take your misinformed whining somewhere else. HTML5 does not have DRM. It merely standardized an API to allow media content to talk to a content decryption module. But there is no actual DRM in there.
You are right, though, that this will have a pretty big impact on the average user's web experience. Namely, it will let them view protected content without requiring proprietary plugins (e.g. Flash, Silverlight, etc). So it's pretty good for the average web user.
>This proposal extends HTMLMediaElement providing APIs to control playback of protected content.
In other words, they're extending HTMLMediaElement to allow Hollywood to put whatever DRM measures in place that they choose. What this WILL NOT do is increase accessibility. Maybe Tim thought it was a good idea, but the old adage is that the road to Hell is paved with good intentions.
The copyright industry are not the good guys, and they will find some way, just as they always have, to make the user experience worse in the interests of their bottom line(s). But who knows? Maybe I'm wrong, or maybe you just have more faith in the industry that was shipping malicious rootkits on music CDs than I do.
EME extends HTMLMediaElement to let it talk to Content Decryption Modules, which must be provided elsewhere. This is just a standardized API for being able to handle protected content. I do not understand all the hyperbole surrounding this.
You're also a bit off your rocker if you're trying to paint "the copyright industry" as some kind of secretive cabal of people who just want to cram DRM on users' computers. Not to mention the fact that sites that don't use this won't even be affected by it. The vast majority of websites out there won't even care, and the ones that do care already have a solution to deal with protected content today, it's just a crappy solution (e.g. requiring Silverlight). Hell, using EME to handle protected content is a far better experience for the user than requiring the installation of a third-party plugin that can has far more capabilities than just decryption content. So if anything, this actually limits the scope of what protected content playback code can do on users' machines.
I never said the copyright cabal was secretive but they absolutely want control over our computers' playback mechanisms.
There was a proposal, I think in the HDDVD spec maybe, wherein the dvd drive itself would be remotely disabled and prevented from playing future content based on the sole discretion of some nebulous 3rd party.
There's the Sony BMG rootkit thing that Mark Russinovich uncovered.
Yes, they absolutely want to cram your computer with DRM and you're a fool for thinking they don't.
They want to control how their content is played back on your computer. That is not the same thing as wanting to arbitrarily stuff arbitrary computers full of arbitrary DRM. Yes, they've gone ridiculously overboard in the past, but DRM is not the goal in and of itself.
