Slashdot Mirror
Luxury Car Hacker To Speak At USENIX Despite Injunction
alphadogg writes "The lead author of a controversial research paper about flaws in luxury car lock systems will deliver a presentation at this month's USENIX Security Symposium even though a UK court ruling (inspired by a Volkswagen complaint) has forced the paper to be pulled from the event's proceedings. USENIX has announced that 'in keeping with its commitment to academic freedom and open access to research,' researcher Roel Verdult will speak at the Aug. 14-16 conference, to be held in Washington, D.C. Verdult and 2 co-authors were recently prohibited by the High Court of Justice in the U.K. from publishing certain portions of their paper, 'Dismantling Megamos Crypto: Wireless Lockpicking a Vehicle Immobilizer.' Among the most sensitive information: Codes for cracking the car security system in Porsches, Audis, etc."
Because if they block the documents, organized crime will never find out.
c++;
US law does not extend outside of the US other, but people have, for example, been arrested for going to places like Thailand and having sex with underage girls and boys. He's still going to be liable to the UK court's decision unless he's never planning to return to his home country.
Fuck the limey court.
Once scientists, even the dim-witted social scientists, get muzzled, the Western Civilization is finished.
Seriously, what's the difference between Ayatollah-mullah fatwas encouraging violence to cartoon drawers, and multinational industrial outfits threatening legal/financial ruins to those who tell truth to power?
Fuck systemd. Fuck Redhat. Fuck Soylent, too. Wait, scratch the last one.
See here's the deal: Just because one person discovers something, it doesn't magically mean that everyone else can figure it out right away. It might be the person who discovered it is pretty clever, and has done a lot of work in that field. So it may well take others quite some time to find it out. If you want to see some examples, look at various military technologies, in particular stealth technology. You might note that that US had working stealth systems long before anyone else.
Now as this relates to security, what it means is that disclosing right away may not be that useful. Perhaps if you give some time for a fix to be implemented, or at least a mitigation, then things could be a little better. Remember with cars it isn't like one can just post a bug fix on a website. All other things aside in terms of what has to be changed, there is pretty extensive testing and certification.
So one can well argue if you've found a flaw in a car you need to notify the manufacturers and give them time to fix it or mitigate it, which may be a good deal of time, rather than running out and telling the world so people know how clever you are.
Like say I discovered that if I pushed on a particular spot in your house, the whole thing would come crashing down on your head. Turns out, said spot is not easy to fix, you can't just go and spend $5 and an hour to do it. It will take a good bit of time and money to fix the problem. Would you like me to let you know, quietly, or would you like me to stick up a poster letting anyone who sees it know, and how that nobody does anything?
Indeed. And normally in cases like this, the researchers alert the people responsible for fixing the problem in good time before publication. In some (many?) cases, the people in charge of the problem doesn't take it seriously, downplaying the risks, or plays the never ending blame-the-contractor game. In that case the only way forward is to threaten to publish the information.
I don't know what happens here, the article never mentions either scenario, but seeing how the people behind the article are serious researchers, I don't think it's very far fetched to guess that they have at least taken some sort of responsible action before publishing the paper. It says that the source code for the crypto has been available since 2009, but hard to know what that means.
c++;
It's a one time pass like an RSA fob. It's synchronized with the cars chip so every time you use it and an exchange happens a net key pair is generated. The key pairs are useless. It's the generator algo you want. Then you'd have to sync up your extra fob with the car.
A fool throws a stone into a well and a thousand sages can not remove it.
Which is the Netherlands. A German company is taking legal action against a Dutch hacker giving a presentation in the US using British law. This is like the poster child for jurisdictional WTFery. Do all laws of all EU member states now apply to every country in the EU? That doesn't sound quite right.
Bio questions? Ask me to start a Q&A journal. Computer analogies available for most topics!
It's not necessary that convoluted. The legal action is taken by the UK-based court on this particular work of the UK-based researcher, working at the University of Birmingham. The original article reads as if the court almost initiated this themselves, due to an ongoing case involving Volkswagen Group. Not sure how that actually holds up.
c++;
Were the tables turned -- if a US injunction prohibited him from publicizing the security flaw -- the US would undoubtedly be leaning on the UK to arrest him as soon as he finished his presentation (or maybe even during it). How much do you want a bet they'll tell the UK to bugger off and deal with it themselves if they are asked to do the same?
See here's the deal: Just because one person discovers something, it doesn't magically mean that everyone else can figure it out right away. It might be the person who discovered it is pretty clever, and has done a lot of work in that field.
History proves you wrong. Usually all it takes is the notion something is possible and vague explanation, or merely advances in other fields that make new discovery feasible. Look up parallel invention.
http://www.kk.org/thetechnium/archives/2009/08/progression_of.php
Who logs in to gdm? Not I, said the duck.
From http://www.bbc.co.uk/news/technology-23487928 :
"The researchers informed the chipmaker nine months before the intended publication - November 2012 - so that measures could be taken. The Dutch government considers six months to be a reasonable notification period for responsible disclosure. The researchers have insisted from the start that the chipmaker inform its own clients."
So essentially they have followed the responsible disclosure protocol but are now being blocked anyway
Top tip: The random number generator in these small battery operated devices is always crap. So the key establishment protocol cannot be secure.
Also, these aren't keypairs in fixed series; there is handshaking to know which keypair in the sequence to use... so you just need to know what a future key is, and keep trying it until it works. That's how they can have multiple keys for the same lock.
See here's the deal: Just because one person discovers something, it doesn't magically mean that everyone else can figure it out right away.
Um, no, that pretty much is the case with computers. Unlike stealth technology (your example), you don't have to have a multi-billion dollar military-industrial partnership to accomplish something clever with computers. You've just got to spend time learning and tinkering. And these security flaws aren't some fundamental-problem-of-physics stuff... it's often just a matter of sniffing out broken/shoddy code, which is pretty much the standard output of the industry. If you think differently, you're probably one of those people who refers to your nephew as a "computer genius" because he got your email working again that one time.
Here's the other deal: companies would rather shut you up than to acknowledge and fix the flaws in their own products. They'd also rather their customers live with the risk. For many companies, it is only the threat of disclosure that makes them invest the time and resources needed to secure their systems. (I'll grant that most software companies seem to have finally accepted this reality and tackle security more heavily, but it seems that the auto companies are still trying to play legal hardball.) And most security researchers (including this one, if other comments I've seen are correct) are willing to give manufactures some advance notice (though it resulted in court troubles in thanks, apparently). But ultimately the utmost professional obligation of a security researcher is to inform the public so they can either protect themselves or force the manufacturer's hand.
-1, Too Many Layers Of Abstraction
The researchers informed the chipmaker
That's the key phrase here. Most likely the chips are not field-reprogrammable. There are no measures to take short of getting new silicon out and recalling the hardware. Knowing the corporate inertia, they'd probably need a year from the date the recall decision was made to implement it and push to the dealers, if they really worked on it like crazy. Fixing crypto where the cost of another mistake may be another recall isn't something you do casually. Presumably some people with suitable theoretical background would need to be contracted and check things out before it hits the fabs. How long would deciding on a recall take I wouldn't know, but presumably not overnight either.
A successful API design takes a mixture of software design and pedagogy.
If you read the actual court judgment, you'll find a slight nuance. The algorithm was invented by Thales. They licenced it to EM. EM and another company called Delphi make immobilizer equipment using it and then sell the kit to Volkswagen and others. The researchers informed EM; EM failed to inform anyone else. VW found out a few weeks before publication and were pissed off.
Which is not to blame the researchers (except perhaps for notifying everyone, rather than simply the maker of the one component they compromised) but it does explain why the judge was fairly scathing: "It may well not be the defendants' fault that Volkswagen were not told earlier, but once the defendants were told about Volkswagen 's concern a responsible academic, concerned with responsible disclosure, would have realised that publication should be delayed, at least for a reasonable period, to allow for discussion with Volkswagen."
And I disagree. The researchers told the company of the product they found vulnerable. This is a security company - they should have measures in place to communicate the flaw up and down. The fact that they did not means they do not take security serious, and they cannot be trusted. There is nothing to fix here - the company has to get out of the security business one way or another.
Should the researchers also listen to any old guy who used a remote locking system on a shed? If I have a VW, can I block the publication because I did not have time to go to the garage yet? Where does it stop? As I see it, VW is just a customer here, and they are at the mercy of the supplier. The supplier can go to court, but VW should stay out of it.