Tesla Fixes Security Bugs After Claims of Model S Hack (reuters.com)

An anonymous reader quotes a report from Reuters: Tesla Motors Inc has rolled out a security patch for its electric cars after Chinese security researchers uncovered vulnerabilities they said allowed them to remotely attack a Tesla Model S sedan. The automaker said that it had patched the bugs in a statement to Reuters on Tuesday, a day after cybersecurity researchers with China'a Tencent Holdings Ltd disclosed their findings on their blog. Tesla said it was able to remedy the bugs uncovered by Tencent using an over-the-air fix to its vehicles, which saved customers the trouble of visiting dealers to obtain the update. Tencent's Keen Security Lab said on its blog that its researchers were able to remotely control some systems on the Tesla S in both driving and parking modes by exploiting the security bugs that were fixed by the automaker. The blog said that Tencent believed its researchers were the first to gain remote control of a Tesla vehicle by hacking into an onboard computer system known as a CAN bus. In a demonstration video, Tencent researchers remotely engaged the brake on a moving Tesla Model S, turned on its windshield wipers and opened the trunk. Tesla said it pushed out an over-the-air update to automatically update software on its vehicles within 10 days of learning about the bugs. It said the attack could only be triggered when a Tesla web browser was in use and the vehicle was close enough to a malicious Wi-Fi hotspot to connect to it. Slashdot reader weedjams adds some commentary: Does no one else think cars + computers + network connectivity = bad?

Comments

Cars?

[ledow]

"Does no one else think cars + computers + network connectivity = bad?"

Does no one else think that phone + computer + network connectivity + radio connectivity + location sensing + chargeable services + .... + ... = bad?

Apparently only a few.

Re:Cars?

[bozzy]

They aren't necessarily intrinsically bad, per se. It's just that people either make mistakes (introduce bugs) or are malicious (abuse it). It's why we can't have nice things.

Re:Cars?

[rudy_wayne]

"Does no one else think cars + computers + network connectivity = bad?"

Whether it's your car, television or phone, it's not bad if done properly. The problem is, nobody gives two shits about doing it properly.

Re:Let me know

[Maritz]

I think they've been on lithium ion for a while now.

Connectng

[fluffernutter]

Connecting a car to anything is just stupid and reckless. It will be a constant battle with hackers. All AI should be on board.

Computers and networks in cars are fine

[sjbe]

Does no one else think cars + computers + network connectivity = bad?

In principle no I do not. Cars have been loaded with computers for quite some time now for all sorts of good reasons. You just don't usually notice them - which is a good thing. As for network connectivity that is fine too. There are all sorts of useful things you can do with network access. Are there downsides? Sure, just like any technology. I haven't seen any showstoppers however. Just problems that will take some time to work through. I think the auto companies are going to struggle for a while to learn to deal with the security issues because they have no experience with them but they'll figure it out eventually. There also are some privacy issues but those too will eventually be sorted out to a reasonable degree.

Actually I think cars without computers are a much worse idea in most cases. Worse performance, worse fuel economy, more dangerous, less features, more maintenance, etc. I'm old enough to remember when cars mostly didn't have computers in them. They're better with computers.

Re:They aren't?!

[fluffernutter]

This goes directly against the whole Autopilot philosophy. In your example, people are expected to be capable of using poison properly and responsibly. If you spread it around and someone doesn't use it properly, well, it's not your fault. Didn't you read the small sign in the corner of the yard? What makes it more interesting is that automation is being pushed on the premise that humans aren't perfect, yet expect perfection from them in other ways? It's a strange way of thinking.