Schwartz On Security: Hack My RideCar security exploits are fast, cheap, and out of control. Why don't automotive manufacturers do more to secure their vehicles?
What do car immobilizers, remote tire pressure gauges, and wireless car-key fobs have in common? Beyond the obvious automotive angle, they've all been hacked using relatively low-cost exploits.
Take car immobilizers and keyless entry systems. A group of Swiss researchers from ETH Zurich recently demonstrated an attack -- dubbed the Autoknacker in Swiss German -- against 10 recent car models from eight different manufacturers, all of which use passive keyless entry and start (PKES). Under this system, a key fob only needs to be physically near the car to deactivate the immobilizer, open the doors, and start the engine.
Here's where attackers come in: By using a pair of wireless antennas, they can rebroadcast the key fob's signal in nanoseconds to fool the car into thinking the driver is present. Furthermore, the antenna picking up the key fob's signal needs to be within only 8 meters (26 feet) of it, making the attack difficult to detect.
The Swiss researchers tested two attack scenarios. First, when a driver is in the supermarket, key fob in his pocket. Second, when the key is left inside a house -- but within range of the antenna -- and the car is parked outside. Using the two-antenna approach, all 10 cars were successfully hacked and driven away.
The total cost for the wireless attack, not including graduate students, ranged from $100 to $1,000, depending on the quality of the hardware used. Interestingly, the cryptography used to secure the fob-to-car communications didn't matter. The researchers were able to remotely fool all 10 cars simply by rebroadcasting the signal.
To block such attacks, the researchers suggest that car manufacturers add security checks to confirm that the key really is located next to the car. "I don't see a way around it," said research lead Srdjan Capkun, an assistant professor of computer science in the system security group at ETH Zurich, quoted in MIT's Technology Review. His group is working on new approaches to help.
While the Swiss researchers' attack doesn't require cracking the encryption keys used to secure the fob-to-car communications, that's also an option. According to New Scientist, security researchers have cracked the keys used by multiple types of key fobs, including the Hitag 2 encryption key from NXP Semiconductors present in many recent cars, although that isn't the only one that's been hacked.
"The proprietary encryption keys used to transmit data between the key fob, receiver, and engine are so poorly implemented on some cars that they are readily cracked," Ari Juels of RSA Labs told New Scientist. Notably, only a few car manufacturers use 128-bit Advanced Encryption Standard (AES) keys. The rest use 40- or 48-bit keys, which security experts now regard as ineffective. Semiconductor makers report that they already have stronger approaches, but it's difficult to convince car manufacturers to use them.
Ditto for wireless car tire pressure gauges. Last year, researchers found that the wireless networks built into many cars in the past three years don't perform authentication or input validation. Researchers thus found it easy to spoof a variety of inputs, such as sensor messages indicating a tire pressure failure. According to the researchers, "we validated this experimentally by triggering tire pressure warning messages in a moving vehicle from a customized software radio attack platform located in a nearby vehicle."
What's the end result of poor automotive information security practices? The next time you finish grocery shopping and find that your car's gone, you can ask: Was it the car crypto -- or lack thereof -- that did you in? Unless a Swiss PhD candidate is lurking in the parking lot, the answer is, probably not. When it comes to bang for buck, there are easier attacks.
But wouldn't it be nice if car manufacturers properly secured their vehicles, so that every new convenience or feature doesn't add another information security attack vector?