Cryptographers Find Fault With Palladium

FrzrBrn writes "Whitfield Diffie and Ronald Rivest raised concerns about Microsoft's Next-Generation Secure Computing Base (formerly Palladium) at the RSA Conference in San Francisco on Monday. They are (naturally) concerned about vendor lock-in and having computers turned against their owners. See the story at EE Times."

Re:Privacy

[neptuneb1]

"I can't wait for the distributed Palladium cracking project!"

You're going to be waiting for a while. With M$'s army of lawyers, any attempt to organize such a project will quickly be shot down by any one of a number of current laws. Let's see how many we can name....

Cryptographers Find Fault With Palladium

[Anonymous Coward]

but due to DMCA laws cannot tell anyone about it, and therefore the faults will never be fixed, because the schmuckos the programmed the damn thing are too damn stuborn, and full of themselves to admit to there being faults in their code, and refuse to fix anything without proof of the faults first.

we now return you to your catch-22 free life . . . no we don't

Re:Privacy

[TeknoDragon]

For every Napster there are a dozen gnutella, hotline, audiogalaxy's... for each of those there's likely to be a clandestine effort to do the same thing.

Besides... we all know there will be someone [nsa.gov] M$ won't be able to stop.

The bit I like

[boy_of_the_hash]

NGSCB also requires secure channels between a keyboard and main memory and between a display interface and a graphics chip and its frame buffer.

Which means it will only work on approved hardware - guess who profits from approving the hardware and drivers? Why would I need a secure framebuffer exactly when I'm already in full control of the code executed on my machine?

It's about who "owns" your ID

[feepcreature]

A central objection from Diffie & Rivest seems to be that under Palladium, Microsoft will own and control your ID - or at least what can interact securely with "your" secure Palladium device.

To understand why this is not a good thing, imagine if a commercial company had the monopoly of passport and driving license production, and were able to prevent you from using the ID they issued to verify who you were except in "microsoft approved" shops and venues (or countries).

IDs and trust systems should be standards based, not proprietary. They should be secure, and openly peer-reviewed or audited. And the ID should be under the control of the person being identified (or at least issued by a "neutral" government body, as passports are now).

But I've just started thinking about this... so I might change my mind some more. Would that make me a bad slashdotter?

what is the fault?

[shird]

From the title, you would think there is some technical flaw in palladium, but the article just goes on about some thing about not having control of your PC etc...

Im not saying there isnt a technical flaw, just /. spreads propaganda through misleading comments.

Suprised MS isn't cyring "conflict of interest"

[pete_wilson]

I'm suprised that Microsoft isn't tyring to cloud the issue by talking about the associations of the persons who gave the talk.

Wittfield Diffie is an engineer at Sun Microsystems, one of the only corporations that can be considered a Microsoft competitor. Ron Rivest is a professor as his day job, but gets quite a bit of cash from RSA, and Microsoft isn't using any of the code that RSA provides (BSAFE, etc) in Paladium, so that's a big chunk of change that won't be coming his way.

We here on slashdot may realize that Rivest and Diffie are actually quite excellent individuals in their field, but these kinds of conflicts of interest are frequently what will be pulled out to counter an argument, rather than working from the facts themselves.

Misleading headline

[BlueFall]

The headline of this story is misleading. Some people disagree philosophically with Palladium's goals, not its technical merits. It just happens that these people are famous cryptographers. At the moment, the technical details seem sparse, so we'll just have to wait until they are released (if ever) to see if the goals that are mentioned are actually met.

Re: Better they find fault with it now,

[Black Parrot]

> And now we're supposed to trust 'Trusted Computing'?

"Trusted Computing" is supposed to fix it where content vendors can trust us.

Or rather, trust our computers.

this just won't fly....

[smd4985]

if foreign governments are having misgivings about using Windows because it is closed source, they surely won't accept Palladium if MS has undue influence and control over the architecture.

Not owning your computer eh?

[scourfish]

It's not much of a change from now: you don't own your copies of windows nor do you own your XBOX

debates? opposition?

[Anonymous Coward]

"We need to understand the full implications of this architecture. This stuff may slip quietly on to people's desktops, but I suspect it will be more a case of a lot of debate," he added.

Rivest said some experts have discussed setting up a forum in technical society for such a debate, but he was unaware of any current moves to do that. Likewise Diffie said he was not aware of any specific alternative to NGSCB in the works at Sun.

I hate to take this stance, but the above says it all. Just like the vast majority of /. that would rather post than write to their representatives, Palladium will simply be buzzworded and adopted by the masses. Regardless of how the technical community kicks and whines, the forces of market domination will likely persevere.

Re:what is the fault?

[Slowping]

From the title, you would think there is some technical flaw in palladium, but the article just goes on about some thing about not having control of your PC etc...

I'd say that the owner not having control of their own keys is a major technical flaw of "trusted computing".

Re:Misleading headline

[wytcld]

Some people disagree philosophically with Palladium's goals, not its technical merits.

How do you separate these two? Having a car you don't hold the key to, but instead have to call some central bureau on your cellphone to unlock wouldn't just be a philosophical problem, but a technical one. It would totally suck technically if your cellphone wouldn't work, for instance - and this vulnerability would be technically more likely than if you carried your own key - a higher rate of failure at car starting. Now philosophically, you may be against always reporting to a central bureau when you'd like to start your car; but technically the scheme still sucks. Same if it's a key to your computer.

Approved hardware

[overshoot]

Why would I need a secure framebuffer exactly when I'm already in full control of the code executed on my machine?

You missed Part Two: you can't get your hardware approved if you don't agree to keep the operational specs under lock & key. So, in order to sell display devices to the monopoly market, they have to be Microsoft-only display devices. Et cetera.

Re:Cryptographers Find Fault With Palladium

[offpath3]

They found fault with the way the computer has more control than the user. They didn't find a crytographic fault in any of the protocols.

Re:This sums it up

[zurab]

From the article: The Microsoft approach "lends itself to market domination..."

Does anyone think Microsoft would have it any other way?

DOJ sues MS for violating U.S. antitrust laws. Courts whole-heartedly agree and rule that MS is guilty. Courts do virtually nothing to protect consumers and tech industry, and literally nothing to punish MS. Courts do not implement any *preventive* measures against MS - as required by the law. MS goes on breaking the same law again and again - nobody pays any attention. MS widely announces its plans (as a marketing campaign) to break the same law again in many-fold worse than before - Palladium - nobody cares.

MS has literally and (seems) legally bribed all - legislative, executive, and judicial - branches of government in order to escape and be exempt from the law, even after it has been convicted of violating it. At some point, the government corruption needs to end, but noone knows how; in the information age where most of the "information" is spoon-fed by corporations that are part of the corruption scheme, the masses will never be on the reform side.

Re:Laws of Robotics?

[archnerd]

The exact laws of robotics are as follows:

1. A robot may not injure a human being, or, through inaction, allow a human being to come to harm.
2. A robot must obey orders given it by human beings except when such orders would conflict with the First law.
3. A robot must protect its own existence as long as such protection does not conflict with the First or Second Law.

Palladium violates all three. A user could be severely inconvenienced by it, it clearly will refuse to obey the user, and it tempts the user to take a sledgehammer to it.

In the Foundation series a "zeroeth law" is introduced which states that a robot must not harm humanity, or, through inaction, allow humanity to come to harm. Palladium screws that up too.

Re:Questions:

[Dave_bsr]

I believe you are incorrect... Last i checked, the way Palladium works is that data is trusted and encryped, not programs... and no doubt, you can run other OS's on your X86 hardware. To make linux on x86 impossible or harder would be retarded for AMD and intel. This is about securing data. Sure, programs are data...but they don't all have to be secure. You can run unsecured code. You can run a whole "insecure" OS, that just happens to be more secure that win_Palladium. BSD, Linux, will be FINE.

Monopoly

[Trevin]

They are (naturally) concerned about vendor lock-in

Isn't this the real reason Microsoft started developing Palladium in the first place?

Re:Not all authors will switch to DRM

[Dr Reducto]

The key word there is "existing". Do you really think Software companies are'nt going to jump on this bandwagon? Microsoft will show them numbers and a graph that points up, and then they will be sold.

Re:The alleged benefit of the CBDTPA, Pd, etc.

[Waffle Iron]

The excuse given for the CBDTPA, which may apply to Pd as well, is that more authors would be willing to publish works in a digital restrictions management system than in a system that grants all fair use rights by default.

Many people throughout history have made great sacrifices to ensure our freedom. Now it seems there are some people willing sell everyone's freedom to use a general-purpose computing device in exchange for a few extra TV shows, video games and pop songs.

I say if the price of freedom is fewer published works, so be it. We're already wallowing in an ocean of media crap anyway; it's not even a big price to pay.

Re:Privacy

[meowsqueak]

In the USA and perhaps a few other countries perhaps - the rest of the world isn't drowning itself in stupid laws quite like the USA is at the moment. Microsoft has a long legal reach but it doesn't extend over the entire planet.

I can imagine 7 years or more down the track, when innovation has been finally eradicated from the US economic landscape, India (for example) will have observed and learned from the USA's mistakes, and become the largest economic superpower on Earth.

Once again, it makes me feel all warm and fuzzy inside to know deep in my heart that no matter how you look at it, I don't live or work in the USA :)

You might be missing a point

[Righteous Indignatio]

In spite of the imagined throngs of doe-eyed deer-in-the-headlights otherwise thoughtless "consumers" out there, it's going to come to pass that Microsoft and their greed will overextend itself. The lock-out we-control-your-security methodology will only work until even the more moronic people have been bitten by it. Perhaps too late for their immediate circumstances, even the most ignorant and go-with-the-flow types will realize they have to leave this Microsoft environment. I believe what we are seeing is two things (a) desperate paranoia-fueled greed and (b) the beginning of the end for anyone so foolish to be so exclusive to the world's computing community. Here on this forum, I keep hearing people talking in little boxes about Intel, Microsoft, AMD, Linux, PCs and all of this shit in this little world we have encased ourselves into. I used to be one of those people. While now I'm working much of my time in Linux (although Windows world stuff still pays some of the bills and mainframes pay the rest) I have gone to a point from being immersed in the Microsoft environments to now being largely outside of them. People? Notice that we are the majority. And we can choose whether or not to be consumer cattle thoughtlessly following the loudest noise. We can choose our own directions. But mental and philosophical freedom is hard work. Not going with the large groups of clueless cattle to slaughter means a lot of effort. If this philosophy of "security" is a bad thing, and I sincerely believe "Palladium" is a very bad thing, don't follow it. Just. Don't. It will have some nice bells and whistles, but recognize a gilded cage and a machine under perpetual remote control and remote authorization for what it is. Don't sit there whining about how Windows 98 or Linux is your favorite OS of choice--please get your egos out of this and start working on some of the deeper principles of your liberty and facility with your own data on your own computers. If it means developing GPL-equivalent hardware, open design microprocessors, and a true open and truly standard machine architecture, done somewhere in the world, then accept this as the direction. Locking people out means locking yourself from them. We have a greedy minority of producers locking out and constricting a vast majority of consumers. Linux demonstrates that we as people can produce, but most of us are in the software or user spheres. People? If they are so intent on locking us out with these obviously evil "security" schemes--let them! But don't let yourself ever be locked in. Linux and OSS is one way to freedom (like Richard Stallman's idea of Freedom as liberty--not lack of cost or price). But perhaps leaving Microsoft, Intel, "Wintel", and going to newer, more open and honest architectures is the way to go. Wintel is rotting and dying. Linux and it's philosophies of openness will succeed because they allow people freedom and the proliferation of new and open idea. Wintel is like the dinosaurs in a sense of being widespread and formidable in the small computer market. This chapter of overreaching greed is the first few pebbles of the beginning of a meteoric shift. Look for freedom and reject this and all attempts to hijack and tyrannize computing.

Re:Laws of Robotics?

[TCaptain]

A Palladium-enabled computer prevents untrusted code from trying to destroy it.

God some people just don't get this...Palladium will NOT stop most of the viruses and worms out there for the simple reason that a virus like code red or melissa or "I love you" does NOT run untrusted code...its a macro run by an application like OUTLOOK...in other words a TRUSTED application.

Palladium is NOT intended to make OUR computers safer from attack, as they are trying to tell you...Palladium exists to give THEM control over OUR hardware...period.

Re:Questions:

[spitzak]

Palladium has absolutely ZERO effect on any end-user security. If the end user has a desire to be secure and has control of the machine Palladium adds NOTHING, NADA, ZILCH. All talk about "security" is a smokescreen.

Palladium cannot stop viruses at all unless all "trusted" programs that could be told to execute the virus instructions actually can't do anything, which would mean the computer is useless. Outlook viruses work by doing things that the "trusted" program Outlook thinks are perfectly benign, the actions are harmful either due to bugs in Outlook or mistakes in it's design. All palladium does is "sign" the bugs in programs and then claim they are "trusted" as though that magically made the bugs go away. It provides no more help than the kernel-mode bit that is already in the hardware and is used by Linux and Windows and does not seem to have stopped viruses on either one of them.

The purpose of Palladium is for Digital Restrictions Management (DRM). There is NO other reason for Palladium. NONE. It's purpose is to make sure that certain programs (everything not written by MicroSoft) does not run on the machine.

The "target" audience is MicroSoft themselves. They are trying to make a machine that is acceptable for playing digital content, with a design that guarantees that alternative operating systems are totally unable to play this content. Far more reliable DRM systems (hardware cards) that would work under Linux are discouraged because of the bogus promises of Palladium.

Re:Laws of Robotics?

[Alsee]

Paladium makes your computer more secure and not less secure. Seems pretty simple to me. Macros aren't trusted code... the next Melissa won't be able to validate the code as trusted and won't be run.

MORON Anonymous Coward.

If that were true then noone would be able to use macros at all unless they subitted it to Microsoft for approval and signing first. And it would cost thousands for evaluation.

Palladium will not prevent macros from running.

Palladium will not prevent you from getting a virus.

Palladium will not prevent a virus from wiping your hard drive.

The only thing Palladium will do is prevent a virus from making a copy of your music files. The virus might be able to STEAL your music files though. The amusing part is that Palladium would then enforce that your copy of the music MUST be deleted when the virus steals it.

They don't care if someone steals your music, they just make sure you can't make copies. You can buy a new copy to replace the stolen one. They are happy just so long as they've made sure you've paid for your copy and you've paid for the copy the thief has. Wonderful system, isn't it? LOL

-

Re:While this is a good start...

[Anonymous Coward]

Here's a go:

Palladium is very similar to what a minidish box or cable box is. For the home user, it's mean to secure a channel to your computer so that digital movies and music can be delivered to your system without you being able to steal it. The difference is that a computer downloading something from the internet can give feedback, whereas my dish network box can't (because I unplugged the phoneline from it).

The reason that this is causing a stir among security experts (and Rivest and Diffie are as expert as it gets. If they say Palladium is insecure, then there'll be hundreds of thousands of people who know in their bones that Palladium is insecure and won't take a second look at it, other than to frustrate themselves with lots of math) is because a computer isn't just a dumb satellite box. Your satellite box stores _no_ information about you. Depending on how you use your computer, it has an enormous amount of information about you on it. Banking information, sexual preferences, medical conditions, credit history, criminal history, and many other potentially embarassing or damaging bits of information are on your computer. For Palladium to work correctly, it'll be able to instantly tie you (as in your name, address, phone number, credit card number, social security number, etc) to any information that accidentally leaks off your computer. An example of this is the referrer tag in your internet browser. If you have javascript turned on, and you go from the goatse [goatse.cx] site to a site that you have to log into, then they can reference that information. It's not a matter of if they can do it, it's a matter of will they do it. Do you want to leave your information up to that?

This is only the tip of the iceburg. What people are really worried about isn't that your computer might leak information, it's that information might be actively taken from your computer. If you stream a song through Palladium, and Palladium thinks you tamper with it, is it going to send your name/address to the media company? Will that company try to extort a fine from you unless you can prove you didn't steal the song? Also, if an application is considered to be secure, are you more likely to dump information into it? If so, what happens when Palladium fails? Microsoft was convicted of being a monopoly for a reason. They don't care about you, they care about your money. They aren't really a company to trust, and the decisions they can make about how Palladium works on your computer are ridiculous.

For businesses, let's assume that Palladium is running on one hundred million computers worldwide (I've done this number bit before). Let's also say that it's flawed in a very minor way, such that 0.1% of the time a trusted application is opened in Palladium, it's marked as untrusted (or unlicensed or stolen or unpaid-for). Assuming each computer only opens one application per day, that's 100,000 daily errors. These errors can't be cleared automatically, because it'd flat be stupid to let errors go without checking on them. That means some call-in center has to resolve 100,000 problems per day. The operators for that service have to be paid, so there'd probably be a priority queue based on who's paid for the premium call in service (I'm not being verbose here because I'm sleepy). If your application errors during a critical presentation or a critical battery of tests, you're just plain fucked (the careful reader will note that this means that Palladium can't be trusted in critical roles, meaning it's already philosophically failed). A computer that might rebel against you is a risk. Furthermore, if there are false-negatives, then there will also be false-positives. Since computers are deterministic and the internet information exchange is so great, once someone find a false-positive, it will be exploitable. Bug free programs don't exist, and Microsoft is worse than usual at churning out bug-free code.

In summation, Palladium can't w

Who owns you?

[0xB00F]

From TCPA / Palladium / NGCSB / TCG Frequently Asked Questions [cam.ac.uk]:

TCPA stands for the Trusted Computing Platform Alliance, an initiative led by Intel. Their stated goal is `a new computing platform for the next century that will provide for improved trust in the PC platform.' Palladium is software that Microsoft says it plans to incorporate in future versions of Windows; it will build on the TCPA hardware, and will add some extra features.

This means that this whole Palladium/TCPA monstrosity requires support from both hardware and software. It is entirely up to the end-user whether or not he wants this. However, senator Fritz Hollings of South Carolina is working on getting a law that will make TCPA mandatory, see here [salon.com]. Until such time that this bill becomes the law:

1. Don't buy the hardware. Unless there is a compelling reason to do so. Well if you are working for the military then go knock yourself out.

2. Don't buy^H^H^H lease/rent/license/WTF the software. There is no compelling reason to do so.

It will only be compelling to use Palladium/TCPA software and hardware only if it becomes illegal not to use it.

Secure computing is not the aim of Palladium/TCPA. Its aim is to provide a way for software peddlers like Microsoft and content pushers like Disney to monitor what you run on your computer and assert control over your computer. In the long run, it will provide them a way to assert control over you.

Secure computing can be achieved through a combination of secure computing practices, secure operating systems running secure applications, and plain-old common sense.

If Intel, Microsoft and their cohorts push through with this stupidity it could spell the end for them. Just think, why in the hell would I want to run this sort of crap? Unless it's mandated by law, there's no reason for me to do so. With the recent slew of news about stupid laws being implemented in the U.S. it's a real possibility.

0xB00F, stands in front of Bill Gates, raises hand, extends middle finger.

Another aspect not mentioned

[GerardM]

Palladium whatever does not mean that the concept of using "software" on only one platform and you have to pay for on another platform is an open and shut case.

The CD's that I have I can play in my stereo, in my car, wherever. With media that is locked to one machine, I CANNOT use said media as I am used to.

Technlology like Palladium should be used EXCLUSIVELY to trusted computing (you know me and i know you..) I am master of my identity, media I buy should work on all my machines. Preventing this is an infringement of my rights and a denial of me as being my own master.

To put it in an scenario; I download (payed for) music, I burn a CD and I will be able to use it in my computer(s), stereo and car.

I am not my computer and my computer is not a customer.

Thanks,
Gerard

Re:Privacy

[Steeltoe]

Do you really think MS is dumb enough to chose a key length that has any chance of being broken anytime soon?

Yes. Does that answer your question?

All they need is the DMCA to stop it from being legitimate. With the DMCA, good security is "not necessary" to keep the masses down, just the law and a police force.

Re:Privacy

[Anonymous Coward]

"You're going to be waiting for a while. With M$'s army of lawyers, any attempt to organize such a project will quickly be shot down by any one of a number of current laws."
An internet project can be based in any other country in the world. Possibly americans cannot contribute, but I don't doubt there'll be enough enthusiasm from the rest of the world to crack in in no time.

Re:Privacy

[Alsee]

M$'s only way to mitigate this would be to force mandatory windows updates

Yes, I have no doubt that mandatory updates will be integral to paladium. Many (most?) palladium programs will only work if they have an active internet connection to the secure servers. This means they can push mandatory updates onto every machine almost instantaneously.

buffer overflow

I think Microsoft is going to come out with some supprisingly solid code for palladium core. They never really cared about bugs before. Now they care and they are going to spend the money and make sacrifices to secure the code. There are tools that can completely eliminate buffer overflows for example. They will keep the core small.

Their code won't be perfect, but you won't be seeing "exploit of the week" in palladium core code. Most bugs will be in palladium applications, and as you said the attacker would run "with the (trusted) credentials of the attacked process". I *think* the system is designed so that access will be isolated to data written by that one application. A bug in your secure e-mail program won't let you crack the DRM on your MP3's.

Between the small bullet proof core, exploit isolation, and the ability to force patches immediately I don't hold much hope in substantially hurting palladium through the usual software exploits. The only way palladium is likely to get a major black eye through this route is if someone holds a virus in reserve until an exploit is found. Then he could release a zero-day virus before they push a patch. A "benevolent" virus could strip the DRM off of files or a malicious virus could just start formatting palladium drives. Someone could write a non-palladium malicious virus, but a palladium version would probably kill off palladium from the public reaction. Hmm, a non-palladium virus could detect palladium and format only palladium machines, lol.

The best attack is the hardware attack to crack each chip individually. The next most likely break is if someone leaks/steals one of the certification keys. There will be dozzens or a few hundred certification keys of varying value. The root key would destroy the entire system. A manufacturer key would force them to void every chip made by that manufacturer in order to save the system. Voiding those chips would instantly kill palladium on thousands of PC in one stroke, the backlash would be staggering.

The least likely attack would be for someone to cryptographicly recover one or more certification keys. Don't hold your breath on this one unless you happen to have some uber-secret quantum computer :)

exploit... to execute one's own applications

No, you can always run your own applications. That is their evil plan, the first hit is free. A "palladium enhanced" computer can do anything a "normal" computer can do. You can write/run your own programs. You can run all old programs. They want make sure that you never have a reason not to get a palladium machine. Once you do have a palladium machine you may end up using a palladium program without even realizing it. Once you do that and you save some data you're locked in. You can never get that data back out of palladium. The more you use it the more you get locked in. The more other people who use palladium the more you'll need to use palladium in order to communicate with them.

-