Massachusetts Sues to Halt Defcon Subway Hacking Talk

According to CNET, "The state of Massachusetts has asked a federal judge for a temporary restraining order preventing three MIT students from giving a presentation on Sunday about hacking smartcards used in the Boston subway system." It'll be interesting to see whether Dutch-style openness or Soviet-style secrecy prevails in Las Vegas. Update: 08/09 20:57 GMT by T : "Too late," says reader Bluey: "Injunction was already granted."

Comments

oh good... let's all bury our heads...

[pha7boy]

rather then make sure they have a techie in attendance so that they may learn something and find a workaround the issue, Boston's lawyers suggested that burying your head in the sand (or, alternatively, in the piles of garbage and crap in Boston) will solve the issue just as well. "As long as we don't let them say it publicly, it does not exist" one Boston official explained the position.

this is why I love government bureaucrats. They tend to be smarter then the average bear.

Re:oh good... let's all bury our heads...

[Mix Master Nixon]

Boston is merely afraid that this information will end up in Lunar hands. Entirely reasonable given that city's sad recent history.

Re:oh good... let's all bury our heads...

[CastrTroy]

What I want to know is how a system like this is even possible. Why should the value available on a smart card actually be something that can be changed by the person holding the card. Shouldn't the card just have an ID, and that ID is tied to an account, which is tied to a person. Maybe put the amount on the card, so the bus doesn't have to call home every time someone steps on a bus, but at least keep all transactions in a database so they can check for fraud after the fact. It seems like the way they have it set up, would be the equivalent of having your bank account balance completely controllable by modifying the information on your bank card. Even retail stores have this figured out so that their gift cards only hold a number, and the actual value on the card is stored in some computer database.

Re:oh good... let's all bury our heads...

[langelgjm]

Maybe put the amount on the card, so the bus doesn't have to call home every time someone steps on a bus, but at least keep all transactions in a database so they can check for fraud after the fact.

I think you hit the nail on the head with this. I don't know about the Charlie card system, but the issue with many transit cards is that it's difficult or impossible for moving vehicles to always be able to check in with the network database to determine the value of an account. So the account value has to be stored on the card.

This is exactly like storing the value of your ATM or gift card on the card itself. But with ATMs and gift cards, the terminal where you use them is always going to have network access (or if it doesn't you probably won't be able to use the card).

Of course, even just storing an account number or identifier on a card doesn't make it fraud-proof. Magstripe cards are trivially easy to re-encode with only a few dollars worth of equipment. Copying these can mean defeating physical access systems, being able to use someone else's gift card balance, or worse.

Re:oh good... let's all bury our heads...

[Jah-Wren Ryel]

Why should the value available on a smart card actually be something that can be changed by the person holding the card. Shouldn't the card just have an ID, and that ID is tied to an account, which is tied to a person.

With a correct implementation - that uses good cryptography - it is quite possible to have secure stored value cards. One upside to stored value cards, especially to slashdot readers, is that they help to protect our right to travel because they can be just as anonymous as cash.

Re:oh good... let's all bury our heads...

[mpe]

What I want to know is how a system like this is even possible. Why should the value available on a smart card actually be something that can be changed by the person holding the card.

Because the people designing these systems don't know what they are doing. This dosn't just apply to RFID systems. There was a case recently involving a magnetic strip card which could be "cloned" by the using nothing more sophisticated than scissors/knife together sticky tape/glue

Shouldn't the card just have an ID, and that ID is tied to an account, which is tied to a person.

Unless it's intended to also use the system to track specific individuals then you don't need any such tying. Just a method of ensuing that every ticket has a unique ID. That only one instance of a ticket with a given ID is in use at any time in the system and that a "never issued ID" or one reported lost/stolen cannot be used.

Maybe put the amount on the card, so the bus doesn't have to call home every time someone steps on a bus, but at least keep all transactions in a database so they can check for fraud after the fact.

A bus might well "call home" periodically anyway, for such things as uploading it's position/CCTV footage/etc at this point it can check the tickets which have recently been used. If it isn't possible to operate a data link all the time.

It seems like the way they have it set up, would be the equivalent of having your bank account balance completely controllable by modifying the information on your bank card.

IIRC at one time it was possible get around withdrawal limits by modifying/cloning cards since they used a read/write area to record this information on the card. So as to enable offline/batch operation of machines.

Even retail stores have this figured out so that their gift cards only hold a number, and the actual value on the card is stored in some computer database.

Probably only as a consequence of being exploited though :)

Re:oh good... let's all bury our heads...

[langelgjm]

That's a pretty weak argument. All you need is a laptop with a cellular data connection. If you really have places where you can't get a cell signal, get the cell company to add a picocell at the bus stops or add a Wi-Fi hot spot. Odds are you won't have to add too many of them in any major metro area.

Well, I'm not the one making the argument, I'm just going by what I see being implemented in transit systems. Storing the value on the card means fast retrieval and processing, and no reliance on a network. What if the data links drops for some reason? What if it takes longer than usual to connect? Transit systems have schedules to keep (ideally!).

Furthermore, it's easy to say "get the cell company to add a picocell at the bus stops", but it's not as if a transit system can simply mandate that it be done. Who's going to pay for it? And at what point does the expense of ensuring reliable network connectivity become greater than simply expecting a certain percentage of fraud? After all, this is a transit system we're talking about, not a bank.

If you have access to somebody else's card, yes. Otherwise, if you are able to steal access, your number space is too small. Use a 256-bit number (or 1024-bit if you're really paranoid) and ensure that new numbers are assigned randomly within that space so that your odds of picking a valid number are remarkably close to zero.

I know. That's why I talked about copying. Plus, given that with things like gift cards, the identifier is often written on the card itself, sometimes you don't even need to have a card reader to get the information. Or, you have security leaks. When I was an undergrad, the University of Maryland inadvertently exposed the ID numbers of the entire university population through its LDAP entries. Those same IDs were used as identifiers on the magstripe cards that gave building access, and dining hall access.

Re:oh good... let's all bury our heads...

[cayenne8]

"Right, because my idea of a perfect society is one where I can't use the damn transit system unless I want to give up any shred of privacy about my destination, travel habits, and location."

Well, that does seem to be the goal of the US govt. at this point. The RealID (national id) alone seems to be a huge step in that direction. They aren't gonna let you travel without one soon...within the US even.

Re:oh good... let's all bury our heads...

[crl620]

MIT's student newspaper put the "banned" slides online: http://www-tech.mit.edu/V128/N30/subway/Defcon_Presentation.pdf

Re:oh good... let's all bury our heads...

[Obasan]

I don't agree with the Massachusetts decision to attempt to stifle the presentation. This was foolish on a number of levels, not the least of which was it will probably help draw far more attention to the hack than it otherwise would have obtained.

That being said, it is perfectly reasonable to not "fix" a system if the cost of the fix is more than the cost of fare evasion. Look - in many cities "evading the fare" is as simple as getting on the bus and choosing not to pay. These systems depend on users for the most part obeying an honor system with periodic random enforcement by transit personnel checking for passes / ticket validation. This is done across Europe and in a number of cities in Canada (not sure about the USA). Why do this? For starters most people aren't jerks, and pay their fares. Second, there will ALWAYS be a way to evade a fare system without massive (expensive) enforcement that would cost far more than the added fare revenue. You would not get on one of the systems where there is no ticket check on entry and then crow about how you evaded the system (or you wouldn't without looking like a complete dork).

It's worth noting that this injunction is not analogous to software companies hiding known exploits in their systems where their customers may suffer the consequences. Boston IS the end user.

Moving people from place to place should always be the highest priority of transit authorities. In general most people are good about paying their fares. Dealing with smalltime one-off thieves is a waste of their resources.

If you use the system without paying, you are a thief and you are doing a tremendous disservice to your fellow citizens.

Frist Amendment

[Mordok-DestroyerOfWo]

Who needs free speech anyway?

Re:Frist Amendment

[thermian]

Who needs free speech anyway?

I can't say.

Re:Frist Amendment

[langelgjm]

What does free speech have to do with releasing software that will help people steal from the transit system?. It sound criminal to me, assisting people to steal.

Right... because clearly that's what the MIT students are trying to do. Help people steal. That was their plan all along...

It couldn't have anything to do with revealing flaws in RFID-based transit card systems that are being increasingly adopted by state and local governments all across the nation, and for that matter, the world. It couldn't have anything to do with shaming a government agency into actually getting on the ball and working with its contractor to improve security of its system. It couldn't have anything to do with plain and simply academic curiosity.

What's it got to do with free speech? Maybe that we think they ought to have the freedom to not only do the work they've done, but talk about it as well?

Re:Frist Amendment

[sabre86]

What does free speech have to do with releasing software that will help people steal from the transit system?. It sound criminal to me, assisting people to steal.

Everything. Perhaps because software, and more relevantly, the presentation, is expression and thus protected under the First Amendment? In a free society where participants are expected to take responsibility for both their own actions and the governance of that society, denying an individual information limits his freedom --knowledge really is power and thus important to freedom -- and destroys his ability to make good governing decisions. For any of us to actually be free, society has to make the fundamental assumption that the average individual will not use the powers given to them to commit criminal acts. You seem to be assuming the opposite. Even if you consider it from a "need to know" point of view (and you shouldn't): both the people who buy into this transportation system and the shareholders of the system, who I understand to be the public, have a right to know the strengths and weaknesses of this system. So they -- we -- the public, have a need to know this information to make the best decisions they can about this system. In fact, we the public have a need to know all things that occur in government, in government contracts and in the public life.

Also, I think you're a bit confused on what "assisting" means. There has to be stealing going on for anybody to assisting in it, and I've seen no evidence that there is. By what I infer your definition of assisting to be: "providing any tool or information used to complete a task" then other things that should sound criminal to you include (but aren't limited to): providing a drunk driver with an alcohol (before he was driving), selling a gun, knife, baseball bat, pencil or anything else to someone who then uses it in a violent crime, teaching anyone any sort of OS or computer security theory (if the students are criminal for providing the information to criminally hack the system, is the professor not criminal for assisting the "criminal" students by providing them with information needed to discover the hack?), etc, etc, etc.

Re:Frist Amendment

[nurb432]

Even if that was the intent to show people how to steal ( which it wasn't ), its still a protected right to talk about it.

Now that said, It wouldn't be protected speech if you ordered people to try it themselves.

Much like its a protected to get up on your soapbox about hating a particular race/whatever and wishing them gone, but it wouldn't be protected if you were organizing a lynching.

I hope you see the difference and why its important to the foundation of freedom in our country.

Ron Rivest

[surmak]

The article mentions that the authorities met with the students and Ron Rivest (e.g. the "R" in the RSA crypto system).

It would be interesting to see what his involvement with this project is.

Re:Ron Rivest

[Anonymous Coward]

He was their professor. Their research was done as a part of a class taught by Rivest.

Too late

[Bluey]

It'll be interesting to see whether Dutch-style openness or Soviet-style secrecy prevails in Las Vegas.

Injuction was already granted. Insert Soviet joke here.

Is MBTA actually going to do anything?

[langelgjm]

Is MBTA actually going to get the card system provider to fix the problem? Because from what I've seen, you'll have a hard time even getting the department and the contractor to admit that the problem exists. And even if they do admit it, is the solution going to be any more than "it's unlikely people will exploit this"?

That sort of attitude seems to be how Maryland feels about its AccuVote TS voting machines. Three independent reviews have all revealed flaws with them, but we're still using them, despite the fact that those flaws essentially mean that the contractor has violated its agreement with the State.

Furthermore, I doubt much criminal activity is going to result from releasing the information. Only a few people are going to have the time and patience to actually follow the exploit through, and if the system is well-designed (though apparently it may not be), modifying card data shouldn't be able to damage or disrupt the system.

Re:Is MBTA actually going to do anything?

[Garse Janacek]

One of the problems is that the MBTA is losing money like crazy, in spite of vastly increased ridership because of gasoline prices. They can't afford to do basic mechanical maintenance and now they have to redo their smart card system too!?

They were somehow able to "afford" the many, many millions of dollars required to install this slow, unreliable, and annoying smart card system. That expense was how they were able to justify the fare increase. I would be fine with an increased fare if it was used to improve service, but instead the service is now significantly worse than before, the smart card machines are terrible (every month I have to wrestle with it to get it to recognize my credit card to buy a pass, and I know others who have the same problem), and they haven't even accomplished the original goal.

And, of course, they voluntarily installed this terrible smart card system even after New York tried installing the same system, and it ended up so terrible that they voluntarily ate the lost money and went with another contractor. I never quite heard the rationale for failing to learn from their mistake...

So, yes, they are losing money like crazy, but my sympathy is limited. They've consistently shown that they don't really know what they're doing.

As for the card vulnerability: it's another demonstration of how worthless the system is, but it hardly matters. Part of the justification for the system was to make sure people paid their fares. It has been a dreadful failure at that, but whatever. The number of people who will go to all the trouble of counterfeiting their MBTA passes is dwarfed by the number that will simply trail someone else through the gates or hop on the green line without paying. This has always been the case. It's not a new or surprising point that secure cryptography cannot prevent social engineering. The fact that it turns out to be insecure cryptography just makes the whole thing more pathetic...

Treat it like the DNS flaw.

[eggman9713]

Just do it the way that they tried to do it in regards to the recent DNS exploits. Tell the affected organization (Boston subway system authority) that there is a problem and you are willing to work with them to fix it. If they refuse, just leave them the information and say they have x number of days to fix it and if they refuse to do anything, you are going to the press, which technically is true since journalists are allowed in limited numbers at Defcon as far as I know. That way you give them the courtesy of warning them in advance, but you aren't needing to completely shut up about it or let the problem lie unfixed. As a white hat, this guy has a moral obligation to help get problems fixed before the black hats find out.

Two problems

[belmolis]

I see two major problems with the application for the order. The first is that it claims that disclosure of how to hack the cards constitutes a danger to the public. How so? All these cards are good for is paying the fare. Hacking them allows people to ride the subway for free. That's petty larceny, not a danger to the public.

The second is that the application asked the court to forbid:

publicly stating or indicating that the security or integrity of the CharlieCard pass, the CharlieTicket pass, or the MBTA's Fare Media systems has been compromised.

There's no conceivable justification for that. Even if there is justification for forbidding disclosure of the details of the hack, stating that there is a problem is certainly constitutionally protected. (It is possible that the court did not include such language in the TRO; this is what Massachusetts asked for, but possibly not what they got. Anybody got a link to the actual TRO?).

What I want to know is...

[strabes]

What I want to know is why Massachusetts is complaining about and interfering with a conference happening in my hometown, Las Vegas.

If this happens,

[nurb432]

Its one more strike against the first amendment and another step down the path of the government deciding what you are allowed to know.

Re:"Congress shall make no law..."

[Wonko the Sane]

Well, this is the State of Massachusetts, not Congress...

They already fixed that loophole

"No State shall make or enforce any law which shall abridge the privileges or immunities of citizens of the United States; nor shall any State deprive any person of life, liberty, or property, without due process of law; nor deny to any person within its jurisdiction the equal protection of the laws."

If I tell you how to hack the DC transit system...

[Anonymous Coward]

If I tell you how to hack the DC transit system right here in this post, will DC issue an injunction to have slashdot remove the post? Let's find out!

In the DC system, you have to scan your card to get into and out of every station. Rather than having standard boarding fares like NY, it actually takes into account where you scanned in and where you scanned out and then deducts the appropriate amount for the fare between those two points at the time you scan out.

But say you leave the same station you entered. Maybe you missed your train and decided to take a cab, or forgot something, or got a call and changed your plans, or just want to rip off the DC transit system. Whatever. You always have to scan a card to get out, and if you scan the same card, it doesn't let you out for free, but charges you a minor fee. I think it was $0.25.

So, say you have a standard commute to work and back every day on the DC transit system:
Go into your point of departure and buy two cards, one with the appropriate fare to your destination. Swipe both of them in.
Ride to your point of departure. Swipe the exact fare card out and throw it away.
Go about your business at your destination. When you return:
Buy a new card and swipe it in.
Ride to your point of origin and Swipe OUT the card you only swiped IN at the same point earlier. You just rode there for $0.25.
The next day, swipe that same card in at the same station. Ride to your point of departure, and swipe out with the card you bought at that point yesterday. Another $0.25 trip.
Always continue to scan in and out at the same station using the same card. Every trip between those stations will be $0.25.

There is no expiration on how much time may pass between swiping in and out of the same station for the minimum fee. There is nothing set up to catch that one card is swiped in and out of the same station every day about 9 hours apart, while another card is swept in and out of another station about 15 hours apart. At least, not unless they've fixed it in the past few years.

Obviously, buy the cards you use for this with cash, not a credit card.

If you really want to be a cheap skate, quadruple your money also. Then all repeat rides in the system will be priced at approximately $0.07 each.