Live data from Hacker News

Internet-connected cars fail privacy and security tests conducted by Mozilla

gizmodo.com

491–500 of 660 posts

Re: Internet-connected cars fail privacy and security tests conducted by Mozilla

#491

Earlier quoted context omitted.

> I assume it's a cost thing. Is it though? It's not like they have to reinvent the button each time. Buttons that last a decade or three have already been designed.

Auto margins are ridiculously thin, and if a manufacturer can trim 17 cents off a car’s manufacturing cost by removing a button, they usually will.

Your post under says margins of 6%? 17 cents over a 6% margin on a $30,000+ purchase would be like McDonalds charging for extra salt on their fries.

I’d guess it’s an ease of design and manufacturing decision when you can eliminate so many buttons so easily.

Re: Internet-connected cars fail privacy and security tests conducted by Mozilla

#492

I don't understand. This seems clearly illegal to have policies like this that claim consent by sitting in the car but the "user" has never been presented the policy in a way to consent to! Is this just a good lawsuit away from being thrown out but no one has done it? Is there some particularly fucked up legal precedent that makes this tenuously legally stable?

> You promise to educate and inform all users and occupants of your Vehicle about the Services and System features and limitations, the terms of the Agreement, including terms concerning data collection and use and privacy, and the Nissan Privacy Policy Similarly, how is this kosher? Do we think anyone, even a single person, has ever followed through on this promise they’ve apparently agreed to in using a Nissan?

In a small informal survey we found that _absolutely nobody_ knows about the new intrusive interior and exterior surveillance from cars.

https://cybershow.uk/media/episodes/watchers_2023-07-01.mp3

It is, to all moral (if not legal) purposes, covert and certainly illegal in the UK on all sorts of grounds that have yet to be challenged in court.

The work is in getting the message out to people.

Re: Internet-connected cars fail privacy and security tests conducted by Mozilla

#493
post #84

The more I read about these things, the more I think I'll be driving my 23 year old Toyota 4Runner until the end of my life

I am still driving my '09 4Runner. Cheap to own/maintain, ridiculously reliable. And no touch screen. Perfection! Anything made after about 2015 feels WAY over-engineered, for my tastes at least.

I totally agree I just feel like there's a sweet spot in the early 2000s where crash safety was better (not the best, but way closer to modern) and traction controls were standard but you didn't have all the spyware. My 2000 4Runner was unfortunately designed in the 1990s which means the doors are super thin as are the roof pillars. Not a deal breaker mind you, it's just the sort of thing that I won't want it until it's too late.

Re: Internet-connected cars fail privacy and security tests conducted by Mozilla

#494

Earlier quoted context omitted.

I’m pretty sure most of the car subreddits are managed by dealership employees who benefit from markups via increased commissions. The fact that they make reference to complaining about price markups as economic illiteracy increases my confidence in this hypothesis. I’m tempted to go make a post about rent seeking, deadweight loss and price collusion in oligopolistic conditions but somehow I feel my economic literacy…

> complaining about price markups as economic illiteracy I mean, I hate dealer markups as much as the next person, but it is true. It's just supply and demand. Dealers wouldn't charge crazy markups if enough car buyers weren't out there paying them; and like another commenter said, in these supply-limited conditions, if price wasn't able to rise to a market-clearing level, you'd have shortages instead, and people wou…

I don’t know if it’s a complaint, more being awe struck by others irrational behavior and how far they’ll go by overpaying. Censoring that observation is a little weird and does seem to favor the dealers grip on a subreddit.

Re: Internet-connected cars fail privacy and security tests conducted by Mozilla

#495
post #475

Earlier quoted context omitted.

We just bought a 2023 Mazda CX-5 Select for $29k. No touchscreens in the whole car! And no dealer markup - we paid slightly under MSRP. But... it is certainly satellite connected (you can manage locks, windows, etc. from your phone, as well as remote start). Wish it was included in this privacy investigation. I'd love to know more about what they do with what they know about our car and how we use it.

You can disable the connected services from the settings. It still has an annoying pop-up every time you start the car to enable them.

More details here, see page 77. PDF warning.

https://www.mazdausa.com/siteassets/pdf/owners-optimized/opt...

Re: Internet-connected cars fail privacy and security tests conducted by Mozilla

#496

Note that this is all based on this source material. https://foundation.mozilla.org/en/privacynotincluded/categor... I haven't dug too much into the methodology, but it seems like it's done based on privacy policies rather than actually looking into the car telemetry traffic. It's also written in a very caaual and sensationalist "omagad" tone that doesn't serve well the seriousness of the topic or findings

Thanks.

I read the original article and conspicuously absent is the actual mechanism of the data collection that is occurring here. Some of the claims seem barely believable.

Is data exfiltrated from the phone, via some kind of data access loophole in Android Auto / Car Play? Are the microphones in vehicles actively listening to conversations and shipping them off for analysis? Or is this all purely hypothetical (i.e., they say they have the right to use it, so clearly they must be trying to acquire it...)?

The only techniques I can imagine that might lead to collection of data such as sexual activity are truly egregious indeed, and although I have incredibly low trust in auto makers such techniques seem like a bridge too far, even for them.

Re: Internet-connected cars fail privacy and security tests conducted by Mozilla

#497

Of all the horrible things go on with privacy these days, this is the one I hate the most. I’m a “car guy”, but not the sort that obsesses over old cars (although I do love 60s and 80s cars). I like new tech, I like the advances in engineering we’ve made in new vehicles, I like EVs even. Nonetheless I’ve been in the market for a new car for months and haven’t bought because it’s hard to find any cars that meet my req…

These markups are insane, and I wonder if anyone actually gets anywhere near paying them. I don’t really see the point since for those markups you can just get a nicer car. Unless car collectors really are that out of touch and flush with cash.

Dealers being able to insist on those huge markups ended a few months back. Overall, US auto dealerships now have more cars in stock than usual. Don't take those markups seriously.

Re: Internet-connected cars fail privacy and security tests conducted by Mozilla

#498

Earlier quoted context omitted.

The touchscreens are really a dealbreaker on new cars. They’re not just prone to quick obsolescence, they’re dangerous.

I hate them so much. Otherwise great cars are completely ruined by touchscreens. What was wrong with tactile knobs you could operate one handed without looking? You’re driving for fuck’s sake. It’s why things like the GR Corolla are nearly miraculous in 2023, it even comes stick shift only.

Fwiw, teslas support pretty much everything via voice. You just press the right button on the wheel and say what you want it to do. I’m not disagreeing with your dislike for touch screens, but Tesla can do literally almost everything hands free using voice if that’s an option.

Re: Internet-connected cars fail privacy and security tests conducted by Mozilla

#499

Earlier quoted context omitted.

That is true, but only in automatic mode. In semi-automatic, everything is quite predictable, no?

Predictable, usually. Lag free? Not in my experience. Most of the time there’s a good quarter to half second between requesting the shift and the transmission acting. For me at least, that lag is very effective at disconnecting me from the experience of driving.

Oh, try a DSG from VW. It's freaking instantaneous. 150ms for the complete operation is about the worst case scenario. I can't shift that fast.

Re: Internet-connected cars fail privacy and security tests conducted by Mozilla

#500

Earlier quoted context omitted.

I hate them so much. Otherwise great cars are completely ruined by touchscreens. What was wrong with tactile knobs you could operate one handed without looking? You’re driving for fuck’s sake. It’s why things like the GR Corolla are nearly miraculous in 2023, it even comes stick shift only.

As an European, where we drive mostly stick, it's funny to observe the newfound US love for stick shifts. It's funny because it happened almost exactly when automatics became good! Good dual clutch gearboxes are amazing, but even "classic" ones like BMW's ZF6 or ZF8 are really close. ZF8 is so good BMW uses it in the M3, instead of a dual clutch. With these options, I'd never go back to stick shift. This after having…

A few years ago I got a 53 Chevrolet pickup with a 3 speed (and a nice BW overdrive attachment). Learned how to drive a manual in that truck and will never buy an automatic again as long as I can choose. Actually operating the vehicle is extremely satisfying and fun. I’m swapping out the C-4 in my 65 mustang for a T-5 as soon as I can spare the cash/time. For me it’s definitely not a matter of practicality but a “vinyl” type of thing. My DD has a CVT but at least emulates shifting with paddles
Post reply on HN