←back to thread

1957 points apokryptein | 3 comments | | HN request time: 0.652s | source
Show context
everdrive ◴[] No.42910717[source]
I'm really happy to see this level of detail and research. So many privacy-related articles either wholly lack in technical skill, or hysterically cannot differentiate between different levels of privacy concerns and risks.

People commonly point to Mozilla's research regarding vehicle's privacy policies. (https://foundation.mozilla.org/en/blog/privacy-nightmare-on-...) But that research only states what the car company's lawyers felt they must include in their privacy policies. These policies imply (and I'm sure, correctly imply) that your conversations will be recorded when you're in the vehicle. But, they never drill down into the real technical details. For instance ..... are car companies recording you the whole time and streaming ALL of your audio from ALL of your driving? Are they just recording you at a random samples? Are they ONLY recording you when you're issuing voice commands, and the lawyers are simply hedging their bets regarding what sort of data _might_ come through accidentally during those instances? Once they record you, where is the data stored, and for how long? Is it sent to 3rd parties, etc? Which of these systems can be disabled, and via what means? Does disabling these systems disable any other functionality of the vehicle, or void its warranty? Lastly, does your insurance shoot up if you have a car without one of these systems? etc ...

The list of questions could go almost indefinitely, and presumably, would vary strongly across manufacturers. So much of the privacy news out there is nothing but scary and often not very substantiated worst case scenarios. Without the details and means to improve privacy, all these stories can do is spread cynicism. I'm really glad to see this level of discourse for the author.

replies(3): >>42911047 #>>42911305 #>>42913968 #
1. AlotOfReading ◴[] No.42911305[source]
Those aren't questions that have fixed answers. The data available is pretty far beyond what I'm personally comfortable with though.

One OEM I'm familiar with had such a policy. My org determined that we needed a statistical reference to compare against within a certain area. Some calls were made to the right people and shortly after we had a (mildly) anonymized map of high precision tracks for every vehicle of that brand within the area over some period.

replies(1): >>42911374 #
2. everdrive ◴[] No.42911374[source]
That’s pretty interesting. What was the purpose of the statistical sample? What did your company want to know precisely?
replies(1): >>42912717 #
3. jon9544hn ◴[] No.42912717[source]
I’m assuming insurance or commercial trucking? Or both?