Tesla withheld data, lied, misdirected police to avoid blame in Autopilot crash | Hacker News Reader

Tesla withheld data, lied, misdirected police to avoid blame in Autopilot crash

electrek.co

519 pointsHamuko273 comments

Tesla must pay portion of $329M damages after fatal Autopilot crash, jury says

https://news.ycombinator.com/item?id=44760573

I've seen people say that this reward seems very high for a single crash and I suspect that the behavior described here is a big part of that.

The fact that Tesla doesn't have a process for making crash data available to investigators is pretty indefensible IMO, given they're retaining that data for their own analysis. Would be one thing if they didn't save the data for privacy reasons, but if they have it, and there's a valid subpoena, they obviously need to hand it over.

For context though, note that this crash occurred because the driver was speeding, using 2019 autopilot (not FSD) on a city street (where it wasn't designed to be used), bending down to pick up a phone he dropped on the floor, and had his foot on the gas overriding the automatic braking: https://electrek.co/2025/08/01/tesla-tsla-is-found-liable-in... The crash itself was certainly not Tesla's fault, so I'm not sure why they were stonewalling. I think there's a good chance this was just plain old incompetence, not malice.

The article claims that the software should have been geo fensed in that area but Tesla failed to do that, that the software should have trigger warnings of collisions but it did not do that. So there were things Tesla wanted to hide.

I don't necessarily disagree, but I personally find these "but you theoretically could have done even more to prevent this"-type arguments to be a little dubious in cases where the harm was caused primarily by operator negligence.

I do like the idea of incentivizing companies to take all reasonable steps to protect people from shooting themselves in the foot, but what counts as "reasonable" is also pretty subjective, and liability for having a different opinion about what's "reasonable" seems to me to be a little capricious.

For example, the system did have a mechanism for reacting to potential collisions. The vehicle operator overrode it by pushing the gas pedal. But the jury still thinks Tesla is still to blame because they didn't also program an obnoxious alarm to go off in that situation? I suppose that might have been helpful in this particular situation. But exactly how far should they legally have to go in order to not be liable for someone else's stupidity?

>I don't necessarily disagree, but I personally find these "but you theoretically could have done even more to prevent this"-type arguments to be a little dubious in cases where the harm was caused primarily by operator negligence.

The article says that soem government agency demanded Tesla to actually geofense the areas Tesla claims their software is incapable to handle. I am not a Tesla owner and did not read the small fonts manual, do Tesla reserve the rights that they might also not sound the alarm when the car is going at speed straight into an other car while a driver is not having the hands on the wheel? sounds bad, the driver is not steering, the car is driving on an area where it is incapable of driving still and it is heading into a obstacle and the alarm is not sounding (still from the article it seemed like this was a glitch that they were trying to hide, and that this was not supposed to happen)

Anyway Tesla was forced to show the data, they did tried to hide it, so even if fanboys will attempt to put the blame `100% on the driver the jurry and Tesla 's actions tell us that the software did not function as adevertised.

> I think there's a good chance this was just plain old incompetence, not malice.

The meme of Hanlon's Razor needs to die. Incompetence from a position of power is malice, period.

As a codendeced variant, yes.

A bit more nuanced version is that incompetence from a position of power is a choice.

That seems contrary to my experience. Large, powerful bureaucracies are often highly incompetent in ways that clearly work against their own interests. If it were merely a choice they wouldn't choose to be incompetent in those ways.

I guess you could go even more nuanced and say sometimes incompetence from a position of power is a choice, and I would agree with that, but now the statement seems so watered down as to be almost meaningless.

I feel like this is getting far too abstract to the point that you’re actively losing sight of a very real, very concrete and very specific set of actions they took which don’t appear to have any credible and innocent motives but also happen to perfectly align with why by all reasonable definitions would be considered malicious.

Large companies in the US, especially at the current moment in history, have huge amounts of power vested in individual executives.

If those executives valued not being incompetent in any specific given way (especially in the ways that harm the many), they have the power to change that. They can say "no, we need to make sure this never happens again."

The fact that they choose not to do that, in so, so many cases, has a variety of causes, but in the end what it fundamentally boils down to is that they choose not to do it.

We were talking about people, not bureaucracies. You've already argued "they were just following orders" on behalf of those who implement. Do you mean now also to excuse those who originate? Or do you really think anyone is going to believe, in 2025, that there is nowhere the buck stops?

"clearly work against their own interests"

Perhaps, but perhaps there is a bigger set of constraints not visible to an outsider which the "buerocrats" are trying to satisfy.

It is also a question about whose interests we are talking about. Interests of individual decision makers can be contrary to the company as a whole.

The point is not to come up with a simple rule that is going to give you the correct answer in every case but to come up with a simple rule that is going to give you the best outcome overall. You need to think about it in game theory terms:

For friends they are unlikely going to be randomly malicious while assuming malice for every mistake is quickly going to ruin your friendship. So Hanlon's razor makes sense.

Corporations on the other hand cannot be assumed to have morals or care about you. You are already fungible to them so assuming malice until proven otherwise is not going to make things worse for you. Meanwhile giving corporations the benefit of the doubt allows the truly malicious ones to take advantage of that who, unlike your friends, don't really have any other feedback loops that keep them honest.

This doesn't acknowledge reality. Tesla has a position of power, but that doesn't mean Tesla is free from incompetence or can ever be free from it.

But Tesla has sufficient power that they do not have the luxury of pleading incompetence when things like this happen.

This is both because such incompetence costs people's lives, and because they have enough money that they could definitely hire more or better people to re-check and add redundant safety features into their products.

The problem is, they do not want any accountability for claiming that their cars are "self-driving", or for any of their other errors or willful endangerment of the public.

Trillion-dollar companies run by egomaniacal billionaires do not need you rushing to your keyboard to make excuses for them.

A corporation can hire people and put processes in place to arbitrarily minimize (or not) the chance of an mistake in areas that matter to them. In this case, they did just that; only the thing being optimized for was “not giving data to the authorities”.

The evidence of this trial does not support an “oopsie poopsie we messed up so sowwy” interpretation of events. Tesla’s paid representatives went out of their way—repeatedly—to lie, mislead, and withhold evidence in order to avoid scrutiny. Fuck them and everyone involved with that.

This comment says more about your relationship with the world than it does about the subject matter

Comments like yours are why we're on the slide into fascism with little fanfare. Oops! We accidentally committed treason. No reason to hold us responsible for our actions!

My relationship with the world is one where I want billionaires and companies with unprecedented amounts of power to be held to account when they act maliciously, negligently, or irresponsibly, and not one where we try to find any reason to excuse their bad behavior.

The real question is why don’t you feel the same?

You are arguing with yourself. I said that Tesla isn't free from incompetence despite its power. You started ranting about billionaires.

We're having different conversations.

>I said that Tesla isn't free from incompetence despite its power.

Ok, and what is your point? Do you have a point or was that just a random observation related to nothing and implying nothing? Because it seems like your implication, given the comment you responded to, it's that Telsa's incompetence is not malicious in it's incompetence. That it is just one of those things that the powerful also have incompetence.

But Telsa decides where to put funding and resources, so if they put funding into covering up and hiding data, and they don't put funding into safety systems and testing that their "autopilot" engages and disengages properly, that is malice.

And again, if that is not the implication of your comment, please just let me know what your intent was, and I will correct myself.

Incompetence can be malice if you are in a position where you are required to be competent - like being licensed to produce cars that share public roads with others.

And in both cases they should be held accountable.

I don't have an excuse why their lawyer did the lawyery stuff, but as far as unlinking of the file being done in software, "destroying evidence", I think the explanation is far more benign.

If you're big on privacy, things like logging incorrect password attempts is a big no-no. We have to "thank" the privacy advocates for that.

How do you think the owner of the car would feel if the file was visible in plain sight to the next owner of the vehicle?

Integrity is for suckers. The system rewards those who play by its rules.

> https://electrek.co/2025/08/01/tesla-tsla-is-found-liable-in...

> Update: Tesla’s lawyers sent us the following comment about the verdict:

> Today’s verdict is wrong and only works to set back automotive safety and jeopardize Tesla’s and the entire industry’s efforts to develop and implement life-saving technology. We plan to appeal given the substantial errors of law and irregularities at trial. Even though this jury found that the driver was overwhelmingly responsible for this tragic accident in 2019, the evidence has always shown that this driver was solely at fault because he was speeding, with his foot on the accelerator – which overrode Autopilot – as he rummaged for his dropped phone without his eyes on the road. To be clear, no car in 2019, and none today, would have prevented this crash. This was never about Autopilot; it was a fiction concocted by plaintiffs’ lawyers blaming the car when the driver – from day one – admitted and accepted responsibility.

---

Personally, I don't understand how people can possibly be happy with such verdicts.

Recently in 2025, DJI got rid of their geofences as well, because it's the operator's responsibility to control their equipment. IIRC, DJI did have support of the FAA in their actions of removing the geofencing limitations. With FAA expressly confirming that geofencing is not mandated.

These sorts of verdicts that blame the manufacturer for operator errors, are exactly why we can't have nice things.

It's why we get WiFi and 5G radios, and boot loaders, that are binary-locked, with no source code availability, and which cannot be used with BSD or Linux easily, and why it's not possible to override anything anywhere anymore.

Even as a pedestrian, I'm glad that Tesla is fighting the good fight here. Because next thing I know, these courts will cause the phone manufacturers to disable your phone if you're walking next to a highway.

I agree. This hurts competent people who want to have responsibility and the freedom that brings.

> This hurts competent people who want to have responsibility and the freedom that brings.

Perhaps, but does it hurt more or less than getting life-changing injuries and your partner killed by a Tesla?

So, we have to ignore the entire safety record for the entire technology just because one operator has failed to follow the instructions?

This is especially the case for something that was in its infancy back in 2019 when this crash happened.

And you know what we have in 2025 because of those restrictions being enforced since then?

In 2025, Tesla's nag drivers so much, for not paying attention to the road, that drivers no longer keep the much safer versions of autopilot engaged at all, when looking for their phones.

Instead, now, because the issue is "fixed", Tesla drivers simply do the same thing what drivers of any other car do in the situation.

They disable autopilot first, and only then stop paying attention to the road, looking for their phone.

How's that safer?

We're precisely less safe because of these regulatory requirements.

(And, add insult to injury, this court is now using the hindsight 20/20, of these warnings subsequently being implemented, as evidence of Tesla's wrongdoing in 2019, at a time before anything like that was thought to be possible? Even though, now that these warnings were implemented, we already have evidence that these nags themselves make everybody less safe, since autopilot is simply turned off when you need to stop paying attention to the road?)

What safety record are we ignoring? Can you please cite some scientifically rigorous and statistically sound data, evidence, and analysis?

Or are you talking about self-published numbers by the company that is proven to withhold, lie, and misdirect in even official police investigations, subpoenas, and trials where it is actively illegal to do so?

Are we talking numbers with a degree of scientific rigor unfit for publication in a middle school science fair, let alone the minimum standard of scientifically rigorous that members of their team had to achieve to get their degrees, yet somehow fail to do when detailing systems that are literally responsible for the life and death of humans?

So where's your unbiased data then?

Where's your data that these nags make everyone safer, when it's widely known that they simply result in people turning off the entire autopilot/FSD when the operator needs to stop paying attention to the road, to avoid the nags and the penalty strikes?

Where's all the news reports about the crashes without the autopilot engaged? If they were as rare as the autopilot ones, surely we'd have seen some of them covered by the media, right? Or are they so rare that not a single one has happened yet, hence, the lack of any reports being available?

You are the one claiming it has a “safety record”.

You are the one claiming “We're precisely less safe because of these regulatory requirements.”

Support your assertion with scientifically rigorous, statistically sound evidence.

And no, your ignorance of safety problems is not evidence of safety despite your attempts to argue as such. That was not a valid argument when the cigarette companies made it and it is not valid now.

There is no car sold in America which could have prevented this accident.

And would the driver’s actions have been different if they had understood that? Was their lack of understanding coincidence, correlated with their Tesla ownership by no fault of Tesla, or deliberately engineered by Tesla’s marketing approach?

You know, I've talked to a whole bunch of people who actually own Tesla's, who actually work in tech, and most of them are completely unaware about any of these autopilot features whatsoever.

Most people are actually very dismissive of autopilot, and are completely misinformed of the benefits / drawbacks / differences of "FSD" versus "Autopilot".

Most are completely unaware of the improvements of v12 or v13, or differences between HW3 or HW4, or which one they have, or that "autopilot" is free, or circumstances under which autopilot can be used etc.

I talked to some guy last year (mid 2024) who was actually paying $199/mo for FSD v12, before the price drop to $99/mo, and swearing how great it was, yet he has never tried the parking feature, even though it's been released several months prior. He's a software engineer. That's just one example.

So, if anything, Tesla's marketing is nowhere near as successful as these naysayers would make you believe. Because the vast majority of Tesla's own customers are actually far behind on autopilot or FSD buy-in, and are NOT aware of the progress.

An average of 100 people die every day in the US due to traffic accidents, many of which would have been prevented by Tesla-like software. You're obsessing about the wrong side of the equation.

Yes, that is exactly Tesla's line, and it is entirely beside the point.

The point is the EXPECTATION set by decades of basically flat-out lying about the capabilities of "Autopilot" and "Full Self Driving".

Hell, just the names alone entirely point to no-limitations full self driving. Plus selling $10k upgrades to use your car as a driverless taxi with the next software upgrade, which years later, has never happened. The fine print is BS; the message screams nothing but capable driving.

Add to that the instant data collection and real-time deleting it at the scene, and Tesla is 100% wrong. (and I used to admire Tesla and aspire to own one)

And this is 100% Tesla's own fault. They did NOT have to market it that way. Had they marketed it as "Advanced Driving Assistant" or "Super Lane-Keeping", or something, and left the data available to the driver, they likely could have won this case easily. "The guy wasn't even looking, we never even implied it could take over all driving, what the hell was he thinking?".

decades?

Musk joined Tesla with the Series B round in 2005. Doesn't seem that way, but it's been 20 years.

He first started talking publicly about "Autopilot" or "Full Self Driving" in 2013, so 1.2 would be referred to as plural decades. (I didn't have the exact number on hand, but knew it was 1+, and used the proper form; you prompted me to lookup the proper number)

First of all, I don't condone the unlinking of snapshot_collision_airbag-deployment.tar after the upload, but, OTOH, I can also understand why something like that would be done, too. (If you were an owner of the car, would you want a subsequent owner to have the records of your crash? This is why I hate all these privacy advocates, because they ruin it for everyone.)

If you've ever worked at any company where security and privacy are taken seriously, you'd be fully aware that things like logging incorrect password attempts is a straight up CVE waiting to happen, even though it's something that a legitimate user might as well want to be happening to find out who's trying to break into their system. Thank the privacy advocates.

But I fail to see who exactly is misled by the marketing, ESPECIALLY given the non-stop negative media attention Tesla has always had.

I mean, it's literally already called FSD Supervised, and previously it's been FSD Beta. How exactly is that not self-explanatory?

But if you already a conclusion, the name is irrelevant. I mean, did you ever look at the Autopilot article in Wikipedia? It's about the aircraft system. Why do not we not FAA complaining to Boeing and Airbus that their airplanes have this super-misleading "autopilot", even though the pilots must still be supervising the tech?

Yikes

You entirely miss the distinction between trained professional and ignorant consumer, "industrial use only" equipment and materials vs everyday consumer goods, things that require technical training and even certification to use properly, vs goods sellable to any consumer, prescription drugs vs otc.

The technical goods carry a much higher risk and must be used with specific training and context. Using them outside those contexts is likely to harm or kill people, and creates legal liability.

In contrast, consumer goods must be engineered to be safe in ORDINARY circumstances with UNTRAINED people using them.

Tesla is trying to paper over those differences and use technical terms in a common environment and unleash tools requiring skilled supervision to prevent death into a consumer environment, for profit. You are either failing to make the distinction or consciously going along with it.

Actual "autopilot" such as the kind actual pilots use does not give the pilot the ability to disregard the operation of their airplane.

How does calling a feature "autopilot" then give consumers the impression that they can completely hand over operation of the car. Driving a car is a serious task and this driver was extremely negligent in executing that task.

The "Autopilot" on planes is a lovely technical distinction. Hardly any non-aviator knows the details, and the common idea is that the plane pilots itself in most phases of non-emergency flight. And indeed, some autopilot functions can even auto-land according to the flight plan. But there is a bit of wiggle-room there to say people should expect to pay attention.

Tesla's "Full Self Driving" leaves no such room. The only exception would be to consider Tesla to be lying.

"Full", adjective Containing all that is normal or possible. "a full pail." Complete in every particular. "a full account."

That's just the first DDG search. Go to ANY other dictionary and show me where "Full" in ordinary American English usage means anything other than complete, no exceptions, etc.

"Full Self Driving" literally means "It FULLY drives itself".

It does NOT mean "You and the automobile co-drive the car", or "Mostly self-driving", or "Sometimes self-driving", or "Self driving until it can't figure it out or fcks up", the latter of which seems the most accurate.

If Tesla had called it ANY of those other things, they would be fine. And would be honest.

But instead, Musk and Tesla decided to lie in plain language. And for that, they lost the case, and will likely lose many others.

For some people, honesty matters. And honesty is not splitting hairs over differences between detailed technical meaning vs colloquial understanding ("autopilot"), or using the end-goal a decade+ away as the official name and description of the feature set ("Full Self Driving"). That is dishonest.

But Tesla doesn't have "Full Self Driving" period, they only have "Full Self-Driving (Supervised)" (and, prior to that, it's been FSD Beta).

You can't just keep ignoring the "Supervised" bit as if it's not there. Just because you think it's a stupid name, doesn't make it a lie. Have you even tried it yourself? I've tried v12, and it's amazing. It does fully self-drive. Why would they call it "mostly" if the idea has always been that it'd be "full" when it's done and out of beta? And as robotaxi shows, it's literally almost there.

I've just tried searching "FSD site:tesla.com" in Google Search, and basically every single result is "Full Self-Driving (Supervised)", with very few exceptions.

Hint: what does the word ‘full’ mean?

"Full" means it drives itself as long as you "Supervise".

Did you use v12 or v13 FSD? I've used v12 last year (which is far behind v13, and yet more behind robotaxi). I'd enable it as soon as I'm out of the garage, and it'd safely drive me right to the destination.

How exactly is that not "Full"? Why would they call it anything else when it can drive itself from point A to point B without any interventions most of the time?

>>Why would they call it anything else when it can drive itself from point A to point B without any interventions most of the time

Why?

Because the FACT is that you must put in the caveat MOST OF THE TIME, or someone is likely to die.

If they were honest they would call it "Supervised Mostly Self Driving". Even "Supervised Self Driving — you mostly supervise, not drive!" would be accurate.

Again, go to any dictionary and find the definition of "Full". Websters:

>>1 containing as much or as many as is possible or normal

>>2a complete especially in detail, number, or duration

>>2b lacking restraint, check, or qualification

>>2c having all distinguishing characteristics : enjoying all authorized rights and privileges

>>2d not lacking in any essential : perfect

The problem is Tesla and you are attempting to literally change the definition of "Full" to mean "Not Full".

This is lying or deceiving yourself, and deceiving others.

Recognize facts and that language actually has meaning, and stop being part of the problem. Or continue watching your hero lose lawsuits.

>>Why would they call it "mostly" if the idea has always been that it'd be "full" when it's done and out of beta?

Because "Mostly..." is the truth, and then when it is actually "Full..." they can come out and announce that fact with great fanfare. and they would have been honest.

Hell, if they simply called it "Supervised Self Driving", it would be honest, and actually match even your glowing description.

But they do not. Your and Tesla's idea that using the added tagline "(Supervised)" as a legal weasel-word does not work either. "Full Self-Driving (Supervised)" is literally an oxymoron. A thing either drives itself fully, or it requires supervision. One contradicts the other.

IIRC, the "(Supervised)" bit was added well after the first fanfare with only "Full Self Driving" alone, when problems started to appear. And the common initials are "FSD".

Even if the reality of the feature set meets your glowing description, the problem is the small percentage of cases where it fails. I'm sure the guy in Florida who was decapitated when his Tesla failed to notice a semi-trailer turning across in front of him was similarly confident, and the same for the guy in California who was impaled on a construction traffic barrier. The problem is that it is NOT FULL, it is only full self driving until it fails.

>>And as robotaxi shows, it's literally almost there.

NO, it shows the exact opposite.

Nearly two months after the much-heralded rollout of (fully) self-driving taxis, Tesla still cannot put a single car on the road for a single meter without a supervising safety driver. Moreover, there have been numerous reported instances of the cars making dangerous errors such as left turns into traffic, etc.

>>basically every single result is "Full Self-Driving (Supervised)", with very few exceptions.

Again, that wording is a literally meaningless oxymoron, containing two mutually contradictory statements ("Full" vs "Supervised"), thus open to whatever interpretation the listener latches onto. Moreover, the emphasis is on the first word — "Full" — which is the lie.

Driving a car is not a "colloquial" activity; it is a technical activity. Splitting hairs about the technical operation of a car is exactly what a competent operator should be doing.

Regardless, this car did not have FSD, it had "autopilot".

Driving a car in the USA with a license requires passing about a 5-minute written test about rules of the road and five minutes of a driving test. It is an almost universal event for residents of the country.

Piloting an aircraft with an autopilot requires a MINIMUM of 1500 hours of instruction and experience as well as multiple levels of certification (VFR, IFR, multi-engine, and specific type certification).

You are seriously trying to claim that these are even remotely similar activities?

Yes, drivers SHOULD split hairs over the technical operation of the vehicle.

Should and Is/Are/Do are NOT the same thing. Particularly when the company founder and prime spokesperson brays about how it will do everything for you and constantly overpromises stuff that won't be available for a decade (if ever) as if it were here already

sheesh

> Piloting an aircraft with an autopilot requires a MINIMUM of 1500 hours of instruction and experience as well as multiple levels of certification (VFR, IFR, multi-engine, and specific type certification).

What? A full private pilot license only requires 35 or 40 hours of flight time (depending on school type); a sport license only requires 20 hours. Airplanes flyable by new pilots of either type very often have autopilots today.

And exactly ZERO of those licenses allow you to get anywhere near an autopilot system. Those licenses are for Visual Flight Rules only. You are not even allowed to fly in conditions that require you to use instruments.

To get to use an autopilot, at MINIMUM is required an instrument rating which is 10+ hours past Private Pilot AND a Commercial Pilot License which is 250 hours minimum, assuming you can find a plane with an autopilot that requires that minimum and included that in your learning program (so hours counted towards both commercial and type certification). Then, you need to be rated on the particular autopilot system.

On top of that, you need to follow RULES about when the autopilot can be engaged, e.g., not below certain altitudes, conditions, etc.

The point stands that the training required just to be able to understand an aircraft autopilot sufficiently to be allowed to use it is a level of serious specialized expertise not encountered in ordinary civilian life.

Taking technical terms of art from areas of specialized expertise and abusing common misunderstandings to apply them to broad marketing and then relying on legal technicalities to try to say "we told you so" when you did no such thing it is just an advanced form of lying. Musk is fooling you too.

I mean, this is just wrong.

Autopilot on a plane does actually drive the plane. You can go as long as hours without any human input requirement. Pilots can eat, go to the bathroom, what have you.

Of course we have two pilots, just in case, but this isn't necessary - some countries are pushing for one pilot because the vast majority of flying is done by the plane.

That doesn't mean that autopilot systems on planes are more sophisticated. It just means that automating a plane is much, much easier than automating a car.

We also have fully autonomous trains.

No one, and I mean NO ONE who has driven 2 days with FSD has the misconception that a Tesla will drive itself under all conditions, 100% of the time. It _is_ "full self driving" in the sense that it fully, by itself, drives me to and from work everyday, to my child's school, to various events around town, _with minimal interventions_.

This is partly because a large portion of the Tesla buyers were more intentional in their purchase (doubly so if they bought FSD). Accidents happen, and FSD surely reduces the frequency, but a change in the marketing (or the NAME lol, the pearl clutching) would not make a gram of difference.

The law is not supposed to make the utilitarian choice of what causes the least net harm to all people, individual rights be damned. That someone died while unchained is not a compelling argument for everyone to be put in irons.

Everyone without a driver's license and liability insurance is in chains with respect to driving a car on on public roadways and it is broadly unanimous opinion that it makes perfect sense. Its a mix of utilitarian grounds, externalities, etc.

You can't build a large dynamite factory in a residential neighborhood either even if you don't intend for it to blow up.

Well yeah, the law isn't "do whatever you want, there are no rules" either. The important thing is getting your license and insurance is purposefully an extremely low bar. Driving is a dangerous activity, someone could hurt or kill themselves or others every time they get behind the wheel, but so long as you have demonstrated the very most basic understanding of what you're doing (ie getting a license), and taken responsibility for the potential consequences of your actions (ie carrying insurance), you're free to take that risk. Note, you still don't have the right to run over pedestrians, but if you do, you will be held liable, not the manufacturer of your car or the state which granted you a license. You would likely lose your license under such a circumstance, but no one else will lose their licenses to mitigate the risk of future pedestrian impacts.

Zoning laws are a complete non sequitur. The issue with building a large dynamite factory in a residential neighborhood is the threat to the people of the neighborhood who not only didn't consent to live near it but specifically chose to live in an area zoned so that such things could not be built. Building a dynamite factory wherever you want is not something you have the innate right to do. That said, you probably can get a permit (assuming you have the proper licenses and insurance) to build a dynamite factory in an appropriately zoned area.

the great thing is that we live in a society controlled by laws and corporations cant get away with testing everything they want on public roads. your freedom or desire for “responsibility” doesn’t negate others’ rights

The right to negligently operate a vehicle?

There's an interesting philosophical debate about the nature of product liability laws. Suppose I'm selling some gadget which when used correctly is safe, and I've taken all reasonable steps possible to try to make it hard for people to use it incorrectly.

Nevertheless people sometimes still manage to use it in an incorrect way that injures their hands such that it will take a year of treatment and physical therapy before they can use their hands again.

Some people view the point of product liability laws is to make those who are to blame for the injury pay. Under that view I'd not be on the hook.

Another point of view is that it should be about who can most efficiently handle dealing with these injuries. Either someone is going to have to pay for the treatment and therapy to enable these people to use their hands again or they are going to probably end up on disability for the rest of their lives which will be paid for by government (and so indirectly by most of the rest of us).

Who should that someone be?

One candidate is the user's health insurance company.

One problem with that is that in the US there are plenty of people without health insurance. They would be able to get free treatment right after the injury at any emergency room if they can't afford to pay for that treatment, but that would only get them to the point they aren't in any more danger. It would not include the follow ups needed to actually restore function, so there is still a good chance they will end up on disability. Also that "free" emergency room treatment will actually be paid for by higher costs for the rest of us.

Even if the user does have insurance that pays for their treatment and therapy, ultimately that money is coming from premiums of the people that use that insurance company.

This health insurance approach then ultimately comes down to socializing the cost among some combination of two broad groups: (1) those who have health insurance, and (2) taxpayers in general.

Another candidate is me, the gadget maker. Make it so I am liable for these injuries regardless of who was at fault. I know exactly how many of these gadgets are out there. If all injury claims from people using them go through me I'll have full data on injury rates and severity.

That puts me in a good position to figure out how much to raise the price of my gadgets to establish and maintain a fund to pay out for the injuries.

This still socializes the costs, but now instead of socializing it across those two broad groups (everyone with health insurance and taxpayers) it is getting socialized across all purchasers of my gadgets.

The people who favor this strict liability approach also argue that I'm the best candidate for this because I'm in the best position to try to reduce injuries. If health insurance companies were the ones dealing with it and they noticed injury rates are going up there isn't really anything they can do about that other than raise premiums to cover it.

If I'm the one dealing with it and notice injury rates are going up I can deal with it the same way--raise prices so my injury fund can cope with the rising injuries. But I also have the option to make changes to the gadgets such as adding extra safety features to reduce injuries. I might come out ahead going that route and then everybody wins.

The article explains that the crash snapshot shows: - hands off wheel - autosteer had the steering wheel despite a geofence flag - no take-over warnings, despite approaching a T intersection at speed

Letting people use autopilot in unsafe conditions is contributory negligence. Given their marketing, that's more than worth 33% of the fault.

That they hid this data tells me everything I need to know about their approach to safety. Although nothing really new considering how publicly deceitful Musk is about his fancy cruise-control.

When you put your foot on the accelerator while on autopilot a warning pops up saying the car will not brake.

From the article, the data Tesla withheld from the court revealed:

* There was no record of a “Take Over Immediately” alert, despite approaching a T-intersection with a stationary vehicle in its path.

* Moore found logs showing Tesla systems were capable of issuing such warnings, but did not in this case.

Everytime you put your foot on the accelerator while the system is engaged a warning pops up saying the car will not brake.

It happens right away and has nothing to do with any other warnings. If you own a Tesla you have seen this warning over and over.

I own a Tesla, and here's my take on the biggest software issue:

Normal consumers don't understand the difference between "Autopilot" and "FSD".

FSD will stop at intersections/lights etc - Autopilot is basically just cruise control and should generally only be used on highways.

They're activated in the same manner (FSD replaces Autopilot if you pay for the upgrade or $99/month subscription), and again for "normal" consumers it's not always entirely clear.

A friend of mine rented a Tesla recently and was in for a surprise when the vehicle did not automatically stop at intersections on Autopilot. He said the previous one he rented had FSD enabled, and he didn't understand the difference.

IMO Tesla just needs to phase out 2019 AP entirely and just give everyone some version of FSD (even if it's limited), or geofence AP to highways only.

In my experience, even most Tesla owners don't really seem to understand the difference between autopilot or FSD.

However, even though Autopilot doesn't obey traffic control devices, it still DOES issue warnings if taking over may be required.

Most Tesla owners I've talked with, are actually completely unaware of the v12 and v13 improvements to FSD, and generally have the car for other reasons than FSD. So, if anything, Tesla is actually quite behind on marketing FSD to the regular folk, even those who are already Tesla owners.

Not in 2019 it didn't.

Why is that so though? Because of false marketing to the degree that is criminal. Elon does have one excuse: Tesla would be bankrupt several times over except for his purposeful criminal lies. Does he actually care about the company? He pumped out OOM more value from Tesla than anyone in history of any company. That is how much he cares about company surviving. Criminal. And too dum* to think of anything innovative to save the company.

Withholding safety-relevant features unless you pay a subscription sounds like something from dystopian fiction, not something that should be allowed in the real world.

The fundamental problem here is that the way it's presented caused the driver to trust it in a fashion he should not have. The jury slammed Tesla for overpromising, and for trying to hide the evidence.

They shouldn't give anything to the cops.

If the driver is 100% liable without autopilot, then they should be held 100% liable with autopilot.

The law should be clear and unambiguous in this regard until we remove the steering wheel entirely.

The penalties for being at fault with auto pilot on should be even higher, since it may as well be just as bad as driving while texting!

No. If they were allowed to just say "We did nothing wrong" to not have to cooperate with an investigation, it would open the door to trivial abuse. It is reasonable to assume they are at least partially to blame. Just like a regular citizen can be arrested when they have a backpack full of money an hour after a bank robbery happened, because it is a reasonable asumption, Tesla should be submitted to an investigation until it becomes clear that the backpack's content is legit.

<<a backpack full of money an hour after a bank robbery happened>

yeah, man, IANAL, but just having money in a backpack isn't probable cause without specific facts tying that backpack to that bank robbery, and I don't think your analogy holds.

I'm not in the least a fan of police overreach, but if a bank gets robbed and you are in the vicinity directly afterwards with a ton of cash in hand, I'd say that consitutes probable cause to at least look into what's going on. I would be angry if in that situation the suspect were let go without investigation.

How are investigators supposed to determine the "100% liable" without access to all available data? In a typical RTC, police will seek to obtain dashcam footage from other vehicles to determine what happened and then determine liability (more likely the insurance companies or courts).

I realize I might have been overzealous in my need to stress that drivers should be held liable for "auto-pilot", when it's possible that there are situations when a vehicle system truly malfunctions / does not disengage even after driver input.

In those situations, the manufacturer should be held 100% liable, and the NTSB / other authorities would need a way to determine that - probably the same way they determine it when other car's systems fail - like, when some cars' accelerator peddle got jammed.

> The fact that Tesla doesn't have a process for making crash data available to investigators

Perhaps hiding the data like this _is_ their process.

> I think there's a good chance this was just plain old incompetence, not malice.

If you or I did this, do you think a judge would care? No. We would be sitting in jail with a significant fine to boot.

The point is that these businesses consider this a "cost of doing business" until someone is actually put in jail.

That’s not the point. The point is if Teslas marketing led the driver to over estimate the car’s capabilities leading to him engaging in reckless behavior. He admitted on the stand that he was acting careless, that autonomous mode required supervision; however, he also admitted that he thought that the car would drive better than a human and intervene when required _based_ on Tesla’s marketing. When you look at Musk’s tweets and the Paint it Black video the jury agreed that it was not an unreasonable belief that was factually _not_ true and found Tesla 33% guilty of the accident.

The fact that Tesla purposely mislead the investigators and hid evidence was why the jury awarded such a large sum.

> The crash itself was certainly not Tesla's fault

Mixing up who is responsible for driving the car is very much Tesla's fault, starting with their dishonest marketing.

If the driver is literally pressing the gas pedal while not looking at the road and overriding Autopilot then it is very much the drivers fault.

As long as there is no criminal liability for people doing this, nothing will change. This is pocket change for a company, rounding error, as Tesla's valuation has gone significantly since this happened in 2019, six years ago.

There is no company for which $329M is "pocket change". This is a huge fine, even for Tesla.

I think OP's point is - the fine is not large enough to impact Tesla's stock price - which is all Tesla cares about.

It also didn't really seem to impact Tesla's decision to keep pushing Full Self Driving and Robotaxi despite it having obvious severe flaws (because Tesla sees this rollout as something holding up its stock price).

Tesla cares about Free Cash Flow. This fine negatively and materially impacts FCF. Tesla cares about this fine.

They obviously care about losing $300M.

Like, if they have a chance, they are going to keep $300M.

That is obvious.

OP's point is - it is not having any impact on Tesla's reckless decisions.

Stock price is #1 on the list of their concerns. FCF is somewhere much lower in the list.

"in the short term, the stock market is a voting machine, in the long term it's a weighing machine"

They definitely care about the FCF impact of this fine and what it says about future liability.

The question is not how much does this cost vs the value of the company; it's how much does this cost vs the cost to just do things the right way. $329 Million is orders of magnitude more than it would cost to implement the additional warnings and geofencing. It's hard to imagine any possible corner that could be cut that would save enough money to justify that sort of risk.

Further, it's one fee for a single accident. Since then there have been 756 more Tesla Autopilot crashes. If each of those got a similar payout, that would be 80% of Tesla's current market cap. Obviously they won't all payout that much, but if on average an autopilot crash cost Tesla $56 Million to handle (settlement + legal expenses + lost sales), that would wipe out the company's profit entirely. There's no possible justification for leaving such a massive liability in place.

This seems pretty dumb of Tesla, as I find it rather moot to the conclusion of fault in the accident. The obstruction of justice is damning.

Autopilot is cruise control. When you understand this, claiming that Tesla is partially at fault here does not match the existing expectations of other driver assistance tech. Just because Tesla has the capability of disabling it doesn't mean they have to.

This all comes down to an interpretation of marketing speak. If you believe "autopilot" is misleading you'd agree with the jury here, if you don't you wouldn't. I'm no lawyer, and don't know the full scope of requirements for autopilot like features, but it seems that Tesla is subject to unfair treatment here given the amount of warnings you have to completely ignore and take no responsibility for. I've never seen such clear warnings on any other car with similar capabilities. I can't help but think there's maybe some politically driven bias here and I say that as a liberal.

Happy to be convinced otherwise. I do drive a Tesla, so there's that.

If this is the 300M jury case 100% they will win in appeals. The driver is clearly responsible for driving and there’s never a moment of doubt about it with Autopilot

Appeal of what and on what grounds? Are you an attorney or are you just making this up?

[flagged]

the username is ironic silly bun

I am making this up

Based on (almost?) all prior cases though.

Note that the driver wasn't found to be fault-free (they got something like two thirds of the blame), so it's unclear why appeals would overturn this.

For what it's worth:

Hsu v. Tesla, Inc. (Los Angeles Superior Court, Case No. 20STCV18473). Autopilot allegedly swerved a Model S into a median on city streets; plaintiff also claimed an unsafe airbag deployment and misrepresentation. Final result: Tesla won (defense verdict). Jury awarded zero damages and found no failure to warn; verdict entered April 21, 2023.

Molander v. Tesla, Inc. (Riverside County Superior Court). Fatal 2019 crash (driver Micah Lee) where plaintiffs said Autopilot suddenly veered off the highway into a palm tree; suit alleged defective Autopilot and failure to warn. Final result: Tesla won (defense verdict). Jury sided with Tesla on October 31, 2023; no plaintiff recovery.

Huang v. Tesla, Inc. (Santa Clara County Superior Court). 2018 Mountain View fatal crash (Walter Huang) while Autopilot was engaged; wrongful death, defect, and failure-to-warn theories. Final result: Settled confidentially in April 2024 before trial.

Estate of Jeremy Banner v. Tesla, Inc. (Palm Beach County; related 4th DCA appeal). 2019 Delray Beach fatal crash where a Model 3 on Autopilot underrode a crossing tractor-trailer; plaintiffs alleged Autopilot was defective and oversold. Appellate development: In Feb. 2025, Florida’s Fourth DCA limited the case by blocking punitive damages that a trial court had allowed. Final result: Settled in July 2025 (confidential) before a compensatory-only trial could proceed

> but it seems that Tesla is subject to unfair treatment here given the amount of warnings you have to completely ignore and take no responsibility for.

Lol is this for real? No amount of warnings can waive away their gross negligence. Also, the warnings are clearly completely meaningless because they result in nothing changing if they are ignored.

> Autopilot is cruise control

You're pointing to "warnings" while simultaneously saying this? Seems a bit lacking in self awareness to think that a warning should muster the day, but calling cruise control "autopilot" is somehow irrelevant?

> I can't help but think there's maybe some politically driven bias here

Look only to yourself, Tesla driver.

> they result in nothing changing if they are ignored.

That’s not true

> Do I still need to pay attention while using Autopilot?

> … Before enabling Autopilot, you must agree to “keep your hands on the steering wheel at all times” and to always “maintain control and responsibility for your vehicle.” Once engaged, Autopilot will also deliver an escalating series of visual and audio warnings, reminding you to place your hands on the wheel if insufficient torque is applied. If you repeatedly ignore these warnings, you will be locked out from using Autopilot during that trip.

> If you repeatedly ignore the inattentive driver warnings, Autosteer will be disengaged for that trip. If you receive several ‘Forced Autopilot Disengagements’ (three times for vehicles without a cabin camera and five times for vehicles with a cabin camera), Autosteer and all features that use Autosteer will be temporarily removed for approximately one week.

https://www.tesla.com/en_gb/support/autopilot

And you don't respond to your own point about it being called autopilot despite it not being an autopilot

>> If you repeatedly ignore the inattentive driver warnings, Autosteer will be disengaged for that trip. If you receive several ‘Forced Autopilot Disengagements’ (three times for vehicles without a cabin camera and five times for vehicles with a cabin camera), Autosteer and all features that use Autosteer will be temporarily removed for approximately one week.

There are videos of people on autopilot without their hands on the wheel...

> And you don't respond to your own point about it being called autopilot despite it not being an autopilot

I don’t follow what you mean here? Are you confusing me with someone else?

> There are videos of people on autopilot without their hands on the wheel...

You can definitely remove your hands momentarily. I’ve seen people apply a weight to the steering wheel to fool it too. Not sure how people defeating the safety features would be Tesla’s fault.

First of all I stated my bias.

What part of how autopilot is marketed do you find to be gross negligence?

I would ask, what is the existing definition of autopilot as defined by the FAA? Who is responsible when autopilot fails? That's the prior art here.

Additionally if NTSB failed to clearly define such definitions and allowments for marketing, is that the fault of Tesla or the governing body?

I'm pretty neurotic about vehicle safety and I still don't think this clearly points to Tesla as being in the wrong with how they market these features. At best it's subjective.

>What part of how autopilot is marketed do you find to be gross negligence?

The fact that it's not an autopilot is a great start.

>I would ask, what is the existing definition of autopilot as defined by the FAA? Who is responsible when autopilot fails? That's the prior art here.

I don't think the FAA defines terms, and prior art is something specific to patents that has no relevance to the worlds of marketing and product safety.

>Additionally if NTSB failed to clearly define such definitions and allowments for marketing, is that the fault of Tesla or the governing body?

NTSB does not approve of marketing nor does it provide such definitions. On what basis do you have to suggest they did any of the sort that Tesla needed their approval?

>>Additionally if NTSB failed to clearly define such definitions and allowments for marketing, is that the fault of Tesla or the governing body?

It's Tesla's. They marketed a product that does not do what they claim it does. The fact that when it does not do those things it can cause (deadly) harm to others, is why they received such a steep adverse judgment.

>I'm pretty neurotic about vehicle safety and I still don't think this clearly points to Tesla as being in the wrong with how they market these features. At best it's subjective.

Who cares how neurotic you think you are? You haven't come across reasonable in this conversation at all.

> At best it's subjective.

It's objectively not autopilot.

The FAA does define how autopilot can and should be used, and so should the the NHTSA (mixed up the transpo acronyms) for ADAS. I suspect the FTC may address false marketing claims if NHTSA does not.

> You haven't come across reasonable in this conversation at all.

This is a discussion. We can disagree. No need to attack me.

> No need to attack me.

I'm not attacking you, it's a direct response to your frequent appeals to yourself as some sort of authority for reason and sensibility in this discussion, when your responses clearly indicate that you are being neither reasonable nor sensible.

> The FAA does define how autopilot can and should be used,

Yeah... in airplanes.

Appeals to myself? I'm stating an opinion on the regulatory uncertainty of automobile safety. Don't be ridiculous.

My problem with all of this is that the regulatory agencies that cover safety in this country did jack shit to prevent this. Blaming Tesla is a scapegoat. The NHTSA and the FTC could have prevented this, but they're just pointing fingers at Tesla. They dropped the ball.

> Appeals to myself? ... > I'm pretty neurotic about vehicle safety

Like, I'm not making it up. So no need to pretend to be daft about it. You each time referred to yourself as someone particularly capable of rendering a reasoned opinion on the matter.

> Blaming Tesla is a scapegoat.

Not really.

>They dropped the ball.

Tesla's actions are incredibly reckless. No other company is behaving like them so it's an absurd proposition to say its the fault of the regulators. It's not a market problem. It's a tesla problem.

> Like, I'm not making it up. So no need to pretend to be daft about it. You each time referred to yourself as someone particularly capable of rendering a reasoned opinion on the matter.

This is childish and does nothing for the discussion. It's not about me. I'm not claiming to be an authority, nor am I using myself as one.

> Tesla's actions are incredibly reckless. No other company is behaving like them so it's an absurd proposition to say its the fault of the regulators. It's not a market problem. It's a tesla problem.

The FTC hasn't once acted on Tesla's marketing. That's tacit approval at this point.

The NHTSA has issued two? recalls on Tesla's ADAS and that wasn't until late 2023.

Relying on lawsuits from crash victims, based completely on unclear or lack of regulations, is far from the ideal situation. It's only a Tesla problem insofar they make up the vast majority of miles driven with this kind of ADAS. The NHTSA just isn't keeping up.

Is Tesla pushing the envelope? Yeah, of course they are. That's by design of the self certifying approach that the NHTSA landed on. Tesla could do a lot more safety testing, but that doesn't mean the regulators haven't dropped the ball.

https://www.washingtonpost.com/transportation/2024/05/04/tes...

>This is childish and does nothing for the discussion. It's not about me. I'm not claiming to be an authority, nor am I using myself as one.

If it's not about you why did you reference yourself as an authority twice? Calling me childish is projecting.

>The FTC hasn't once acted on Tesla's marketing. That's tacit approval at this point.

Asinine logic.

>Relying on lawsuits from crash victims, based completely on unclear or lack of regulations, is far from the ideal situation. It's only a Tesla problem insofar they make up the vast majority of miles driven with this kind of ADAS. The NHTSA just isn't keeping up.

Who is relying? Tesla is doing what they want. No one is forcing them to sell this dreck.

Continuing to focus on me does nothing to substantiate your position.

> Asinine logic.

Do you want to make an argument, or is your goal just to dunk with these low effort responses and waste my time?

Multiple automakers are making similar marketing claims as Tesla as their level 2 classified technology improves. Tesla is not even remotely alone in any of this. They're just ahead. Meanwhile FTC has done nothing on this front for a decade.

Is a decade not long enough to take action? What's acceptable in your opinion?

> Who is relying?

Everyone that cares about autonomous vehicle safety and thinks Tesla is not doing enough to protect drivers. You'll find plenty of lawyers stating that tort law is the only way to settle ADAS cases right now.

In this particular 2019 crash, the jury found that Tesla had a defect for not disabling autopilot in this area. This is something that NTSB specifically called on the NHTSA to define better in 2017 prior to this fatality. In 2020, after this traffic fatality, the NHTSA responded for the first time saying this would be impractical to impossible to do and doubling down that the driver is responsible, to which the NTSB calls out the NHTSA for not doing enough to prevent these types of fatalities with Tesla.

https://data.ntsb.gov/carol-main-public/sr-details/H-17-038

So we have a situation where Tesla and every other automaker is operating under the ill defined autonomous safety regulations of the NHTSA, a silent FTC, and a safety review board claiming that NHTSA isn't doing enough prior to and after this accident.

Dunk on Tesla all you want, but that's not going do anything to further road safety. I suspect that Tesla will lean heavily on FTC inaction and NHTSA's 2020 response in their appeal. Rightly so given the jury is contradicting NHTSA's own statement on liability for level 2 system. Hence the use tort law.

>Continuing to focus on me does nothing to substantiate your position.

Lol, it's not my position. It's the law. They were held liable. You kept making the focus on yourself, bringing yourself up for no reason, and then even more perplexingly, denying that you did it.

>Multiple automakers are making similar marketing claims as Tesla as their level 2 classified technology improves. Tesla is not even remotely alone in any of this. They're just ahead. Meanwhile FTC has done nothing on this front for a decade.

I haven't seen that and no one seems to be close to what Tesla outright implies with "autopilot".

>Is a decade not long enough to take action? What's acceptable in your opinion?

The fact that regulators should take action is not incongruous with the fact that Tesla is reckless and negligent. Both can be true. No one forced Tesla to do anything so no one else is responsible for their behavior.

>Everyone that cares about autonomous vehicle safety and thinks Tesla is not doing enough to protect drivers. You'll find plenty of lawyers stating that tort law is the only way to settle ADAS cases right now.

Jeez. I guess it's everyone else's fault but Tesla's when it's clear that everyone else thinks Tesla is behaving poorly.

>Dunk on Tesla all you want, but that's not going do anything to further road safety.

Yawn. This comment thread isn't doing anything to "further road safety" how obnoxious of you.

Saying Warnings are meaningless because they can be ignored would literally flip the entire legal system on its head. That is literally an insane way to think about things.

Not all warnings are equal. Are you a negligence or products liability attorney? You think that a warning that is easily ignored and moved past to do the same thing anyway is sufficient to obviate liability? I don't think the law agrees with you there. Not sure what you think is so insane about that.

I might challenge with "autopilot is cruise control." To me, Tesla is marketing the feature much differently. Either way, looking up the definitions of each:

"Auto Pilot: a device for keeping an aircraft or other vehicle on a set course without the intervention of the pilot."

"Cruise Control: an electronic device in a motor vehicle that can be switched on to maintain a selected constant speed without the use of the accelerator."

It IS fancy cruise control.

That is not how it’s marketed at all.

In both cases, they are driver assistance. A pilot is responsible and must monitor an autopilot system in a plane. We license drivers and pilots and the responsibility is placed on them to understand the technology before using it and putting themselves and others at risk.

Would Boeing or John Deere be responsible for marketing language or just the instruction manual. We know the latter is true. It's there any evidence of the former? Intuitively I would say it's unlikely we'd blame Boeing if a pilot was mislead by marketing materials. Maybe that has happened but I haven't found anything of that sort (please share if aware).

Would Boeing or John Deere be responsible for marketing language or just the instruction manual. We know the latter is true

Actually, the former is true. Courts and juries have repeatedly held that companies can be held responsible for marketing language. They are also responsible for the contents of their instruction manual. If there are inconsistencies with the marketing language it will be held against the company because users aren't expected to be able to reconcile the inconsistencies; that's the company's job. Thus, it's irrelevant that the small print in the instruction manual says something completely different from what all the marketing (and the CEO himself) says.

The "autopilot is limited" argument would have worked 20 years ago. It doesn't today. Modern autopilots are capable of maintaining speed, heading, takeoff, and landing so they're not just pilot assistance. They're literally fully capable of handling the flight from start to finish. Thus, the constant refrain that "autopilot in cars is just like autopilot in planes" actually supports the case against Tesla.

> They're literally fully capable of handling the flight from start to finish.

I find this to be a bit of a rosy take on things.

Autopilots don't take off (which is why Airbus' ATTOL project was a notable thing when an A350 took off "autonomously" [1]). They don't handle ATC (with the tenuously arguable exception of things like Garmin's Autoland), or handle TCAS on what I'd say is a majority of airliners.

Autopilot on planes is still quite "dumb".

1- https://www.airbus.com/en/newsroom/press-releases/2020-01-ai...

> Thus, the constant refrain that "autopilot in cars is just like autopilot in planes" actually supports the case against Tesla.

Just like, holds a lot of weight. I'm saying autopilot has a meaning in the world of aircraft and the FAA has some guidance on how it's used. They still place all responsibility on the pilot. So in that sense they are similar.

It's not that I think automakers shouldn't be liable for misleading marketing, it's that in this case I don't think the argument is strong.

> Thus, it's irrelevant that the small print

The driver has to agree to understanding how it works before using the feature. In the manual it's called out in the same way my Subaru calls out warnings for eyesight. In the model S 2019 manual, the car in this accident, it's clearly labeled with multiple warnings and symbols. Saying it's small print is disengenous. Half of the manual in that section is warnings about the limitations of the driver assistance tech.

The fact that it's in in the instruction manual is irrelevant because consumers are not required to read the instruction manual before using the car. And either way, the manual conflicts with the marketing of the car, and with the CEO's many statements saying that the car could drive itself.

A consumer is not expected, nor required, to resolve this conflict that the company created for itself through its own choice of conflicting language. Tesla was able to get away with it longer than expected, but now the floodgates are open to reason again.

> consumers are not required to read the instruction manual before using the car.

Indeed, and perhaps this is part of the problem. A reasonable person would find that through licensure, there is a expectation that you know how to operate and take responsibility for the operation of the death machine you step into.

If Tesla's marketing is so dangerous why hasn't the FTC acted even once? FTC has been pinged by the NHTSA and Congress. At least the first time was before this accident. It took years for NHTSA to implement the nag. NHTSA could have recalled autopilot before this accident happened.

Tesla did not get away with anything. The agencies failed to address it in a self certification model set by them. It's their job to ensure drivers are safe. Meanwhile Germany and the California DMV did do something. If Tesla is to blame, so are the FTC and NHTSA.

The difference is in the sheer amount of training pilots have to go through, and the regulations that they, and their employers, are required to follow. This is tremendously different from a car that throws up a couple of warnings that can be quickly and passively acknowledged prior to your using "autopilot".

You can't passively acknowledge it. It requires actively agreeing that you understand prior to use. The dialog in one version makes reference to autopilot in the pilot definition.

Maybe when this accident happened it was different, but as far as I know it's always been behind a confirmation dialog.

To operate a motor vehicle in the US, you must be licensed. That surely holds some weight here.

While technically an autopilot might sometimes be as simple as maintaining a heading, the actual practical consequence is quite different for a boat or an aircraft than for a car. There is simply not much to crash into when you're in the air or open water. The result is that a much simpler mechanism is required to achieve the same result for the pilot.

When I worked on unmanned vehicles, you could have one operator control multiple speedboats because you typically had minutes to avoid collisions. Splitting attention would not be feasible with a car on cruise control, because you are never more than a few seconds away from crashing into something solid.

You are right, but unfortunately you are the least useful right, which is technical right.

That is definitely what auto pilot means in the aeronautical and maritime sphere.

But a lot of the general public has a murky understanding of how an auto pilot on a ship or a plane works. So for a lot, probably the majority of them. They will look at the meaning of those two words and land on that auto pilot, means automatic pilot. Which basically ends up beeing self driving.

Sure in a perfect world, they would look up what the term means in the sphere they do not know, and use it correctly, but that is not the world we live in. We do not get the general public, we want, but we have to live with the one we got.

You are also merely technically right. It would require an intentional suspension of one's theory of mind to not recognize the extent to how Tesla's own marketing of its products, and its determination to hide pertinent information in cases like this, is intended to perpetuate the popular misconception of its capabilities.

> ...we have to live with the [the world] we got.

There was nothing inevitable in how we reached this situation, and no reason to let it continue.

> "Auto Pilot: a device for keeping an aircraft or other vehicle on a set course without the intervention of the pilot."

All an auto pilot on an aircraft does is keep the plane flying in a straight line at a constant speed. It mostly doesn't do obstacle avoidance, or really anything else. Yes, you don't need intervention of the pilot, because it turns out going in a straight line in an airplane is pretty hard to screw up.

From that standard at least, modern cruise controls are more capable than airplane auto pilots. There is a widespread belief on HN, however, that people are generally very dumb and will mistake autopilot for something more like FSD.

    There is a widespread belief on HN, however, that people are generally very dumb and will mistake autopilot for something more like FSD.

I think the error here is that you're underestimating just how rare accidents are. Let's imagine there's some monstrously dangerous feature X that results in fatal collisions 10% of the time when misused. If we assume a person either uses it correctly or consistently misuses it, how many people (1 in N) need to be misusers to double the US fatality numbers?

You only need about 1 misuser in every 500-2,000 drivers, depending on how you do the numbers. Now obviously autopilot isn't as dangerous as our hypothetical feature X here, but do you think it's reasonable to argue that a small fraction of a percent of autopilot users might be misled about its capabilities by the name? I think that's a long way from saying "people are generally very dumb".

I don’t think they were misled by the name. I dislike musk and Tesla, but the use of autopilot to describe their cruise control is one of those things where they are being fairly pedantic, and, out of character for them, actually technically correct. Anyone who knows what an autopilot does won’t be misled, anyone whose only experience with the term is from the 1979 movie Airplane! (Otto Pilot anyone?) also isn’t going to be misled. Then what we basically have left are HN pedants who themselves personally got the term wrong and didn’t do any due diligence.

My question is whether you believe the term is so wildly obvious that questioning whether a tiny fraction of a percent of drivers misunderstand it is completely unreasonable. It doesn't rely on Tesla having misused the term at all and for the record, I don't think they are.

But, I don't think it's unreasonable that some of the 5% of US adults who have never been on a plane might not understand what autopilot is in aviation. I don't think it's likely that the 8.4% of US adults who score below the lowest measurable level of PIAAC literacy have a good understanding of the warning messages when you enable Tesla's L2 features, or are digging through the owner's manual to understand them. It seems unlikely that the 3% of adults with <70% IQs are reasoning out the limitations of the system from technical definitions. Hopefully the idea is obvious here. You only need one person out of thousands to make a massively dangerous system. I don't think it's an obviously ridiculous argument that one person out of thousands doesn't fully understand and consider the complicated limitations of such a system.

"it's never the crime... its the cover up". So in this case, they are kinda screwed.

I've owned two Tesla's ( now a Rivian/Porsche EV owner). Hands down Tesla has the best cruise control technology in the market. There-in lies the problem. Musk constantly markets this as self driving. It is NOT. Not yet at least. His mouth is way way way ahead of his tech.

Heck, stopping for a red light is a "feature", where the car is perfectly capable of recognizing and doing so. This alone should warrant an investigation and one that i completely, as a highly technical user, fell for when i first got my model 7 delivered... Ran thru a red light trying out auto pilot for the first time.

I'm honestly surprised there are not more of these lawsuits. I think there's a misinterpretation of the law by those defending Tesla. The system has a lot of legalese safe-guards and warnings. But the MARKETING is off. WAY OFF. and yes, users listen to marketing first.

and that ABSOLUTELY counts in a court of law. You folks would also complain around obtuse EULA, and while this isn't completely apples to apples here, Tesla absolutely engages in dangerous marketing speak around "auto pilot". Eliciting a level of trust for drives that isn't there, and they should not be encouraging.

So sorry, this isn't a political thing ( and yes, disclaimer, also a liberal).

Signed... former Tesla owner waiting for "right around the corner" self driving since 2019...

> ABSOLUTELY counts in a court of law

Are there clear guidelines set for labeling and marketing of these features? If not, I'm not sure how you can argue such. If it was so clearly wrong it should have been outlined by regulation, no?

> given the amount of warnings you have to completely ignore and take no responsibility for.

The article says no warnings were issued before the crash.

So which warning did the driver miss?

The one you accept when you first turn it on. And the numerous ones you ignored/neglected to read when using features without understanding them.

This is the responsibility of a licensed driver. I don't know how a Mercedes works, but if I crash one because I misused a feature clearly outlined in their user manual, Mercedes is not at fault for my negligence.

If the feature you misused wasn't a part of your driver's ed class/driver's license test, and was dangerous enough to cause a crash if used improperly, perhaps Mercedes is at fault (to whatever degree) because they didn't do enough to ensure that drivers knew how to use it. Yes, technically, the driver may be at fault because, well... they're the driver, but this isn't something that is "either/or" - both can be at fault.

Drivers need to be paying attention, but is it not possible that Tesla could also do more to make things clear?

The FAA sets guidelines for autopilot, should the NHTSA not do so for ADAS and enforce that on automakers to keep everyone safe? What that looks like I'm no so sure.

I do think that needs to happen. That's pretty central to my point. Holding Tesla liable when the regulators haven't done so, seems like a unfair judgment even if Tesla and so many other automakers can and should do better. But liability should be clearer than a disagreement over the word autopilot in marketing materials.

I'm reminded of Vitamin Water...

the Center for Science in the Public Interest filed a class-action lawsuit

The suit alleges that the marketing of the drink as a "healthful alternative" to soda is deceptive and in violation of Food and Drug Administration guidelines.

Coca-Cola dismissed the allegations as "ridiculous," on the grounds that "no consumer could reasonably be misled into thinking Vitaminwater was a healthy beverage"

Interesting case but I'm not sure it's apples to apples.

One, you don't need a license to buy a non alcoholic beverage. Two, while the FDA has clear guidelines around marketing and labeling, I'm not aware of any regulatory body having clear guidelines around driver assistance marketing. If they did it wouldn't be controversial.

The analogy is that marketing knows their claims are BS, but the people/courts may disagree

Tesla's not being treated unfairly. It advertised Autopilot as having more capabilities than it actually did. Tesla used to sell Autopilot as fully autonomous. ("The driver is only there for legal reasons.")

And it didn't warn users about this lack of capabilities until it was forced to do so. Those warnings you're talking about were added after this accident occurred as part of a mandated recall during the Biden administration.

> Those warnings you're talking about were added after this accident occurred as part of a mandated recall during the Biden administration.

If that's the case, this is certainly a stronger argument. I thought autosteer and FSD always had this dialog. As far as I know these dialogs go back 10 years and this was April 2019.

Even still find retroactive punishment of this to be dubious. If Tesla is liable to some degree so should the NHTSA, to the extent that anyone who makes the rules can be, for not defining this well enough to protect drivers.

If Tesla is liable to some degree so should the NHTSA,

That's ridiculous. Tesla chose to make dangerous claims that resulted in the loss of dozens of lives. Tesla alone should be liable for this, not the regulator that eventually forced them to add the disclaimers so that consumers would have at least some modicrum of notice that Tesla's advertising was actually just a package of lies.

The NTSB says otherwise.

https://data.ntsb.gov/carol-main-public/sr-details/H-17-038

We remain concerned that NHTSA has taken no action to address the intent of Safety Recommendation H-17-38, which, if implemented following the Williston, Florida, crash, may have prevented the similar crash in Delray Beach, Florida, in 2019. Level 2 systems have limited operational design domains; however, manufacturers like Tesla continue to permit automated vehicle operations in operating conditions for which they were not designed or appropriate. Because of NHTSA’s inaction toward addressing this recommendation, the risk of future crashes remains heightened in challenging environments, such as nonlimited-access roadways, intersections with cross-path intrusions, and roadways with high bicycle and pedestrian traffic, to name a few. As NHTSA emphasized in its response to this recommendation, drivers are expected to remain fully engaged in the driving task when using a Level 2 system; however, if drivers’ use of a Level 2 system results in their prolonged disengagement from the driving task—as we have seen in four crashes that we have investigated—it cannot be stated that these systems are being used as intended. This unintended use indicates a fault with the system design, and not something that can be declared the driver’s sole responsibility. We urge you to reconsider your approach to Safety Recommendation H 17-38, which remains classified OPEN-- UNACCEPTABLE RESPONSE.

Do you think Tesla spends more time and money on making their warnings convincing, or making their marketing convincing? If a person is hearing two conflicting messages from the same group of people, they'll have to pick one, and it shouldn't be surprising if they choose to believe the one that they heard first and that was designed by professionals to be persuasive.

In other words, if you bought the car because you kept hearing the company say "this thing drives itself", you're probably going to believe that over the same company putting a "keep your eyes on the road" popup on the screen.

Of course other companies have warnings that people ignore, but they don't have extremely successful marketing campaigns that encourage people to ignore those warnings. That's the difference here.

Tesla infamously doesn't have a marketing team, so that one should be easy to answer unless they rehired for that. Not sure on the latest there.

When you get your Tesla and attempt to turn on the features described, it has a dialog with a warning and you have to agree to understanding before proceeding. If you choose to ignore that, and all the other clearly marked and accessible warnings in the manual and not learn how to operate it, is that not on you the licensed driver? Isn't that what regulations and licensure are for?

I'm very much in favor of consumer protections around marketing, but in this case there were none that were clearly defined to my knowledge.

Yes of course it is on you, that's why the driver in this case was found mostly responsible. The question is whether Tesla bears any additional responsibility, not whether they are solely at fault, which they obviously aren't.

> Tesla infamously doesn't have a marketing team

Come on, obviously they do marketing. "We don't have a marketing team" is itself marketing. Here's their most popular YouTube ad, for example: https://www.youtube.com/watch?v=tlThdr3O5Qo

That video, which is called "Full Self-Driving" even though it came out when Autopilot was the only capability the cars had, was coincidentally released 3 days before the crash we're discussing. Do you think an ad like that, which simply shows the car driving itself and links to a page about Autopilot in the description, might lead someone to believe that Autopilot will do what the video shows? Again, remember that there was no separate FSD product at the time.

> Come on, obviously they do marketing.

Of course they do. You questioned time and money spent compared to warning clarity and the manual. Having no team for marketing implies little time and money spent. I would say they've spent far more money trying to make it safe and clear in the user manual than trying to mislead customers in a few tweets and a video.

> might lead someone to believe that Autopilot will do what the video shows?

It does do what it shows? The driver is attentive not placing their foot on the pedal and not taking their eyes off the road. I believe that is exactly how FSD operated until forced to implement the nag 4? years after this video was made.

Afaik this is the video congress sent to the FTC, under the awesome Lina Khan, and they chose to do nothing about it. I agree Tesla should do better, but that's different than finding them liable for wrongful death in this specific case.

Germany did something about it. California DMV did something about it. If anything it seems the NHTSA and FTC dropped the ball. That seems to be in Tesla's favor legally.

> Autopilot is cruise control. When you understand this, claiming that Tesla is partially at fault here does not match the existing expectations of other driver assistance tech.

The problem is for several years they actively targeted a customer base incapable of understanding the limitations of the mis-named system they advertised. (Those customers capable of understanding it were more likely to buy vehicles from brands who advertised more honestly.) While the current approach of targeting Nazi and friend-of-Nazi customers might eventually change the story (with its own risks and downsides, one imagines), for the time being it seems reasonable that Tesla bear some responsibility for the unsafe customer confusion they actively courted.

> This all comes down to an interpretation of marketing speak. If you believe "autopilot" is misleading you'd agree with the jury here, if you don't you wouldn't. I'm no lawyer, and don't know the full scope of requirements for autopilot like features, but it seems that Tesla is subject to unfair treatment here given the amount of warnings you have to completely ignore and take no responsibility for. I've never seen such clear warnings on any other car with similar capabilities. I can't help but think there's maybe some politically driven bias here and I say that as a liberal.

And that's exactly why the law is supposed to have a Reasonable Person Standard.

https://en.wikipedia.org/wiki/Reasonable_person

When the majority of Tesla's owners are completely unaware of the viability of autopilot even in 2025, how exactly does it make any sense to blame the marketing when someone was so entrusting in the unproven technology back in 2019? Especially given so many reports of so many people being saved by said technology in other circumstances?

I imagine these things will get better when courts would not be able to find jurors that are unfamiliar with the attention-monitoring nags that Tesla's are famous for.

Apologies if this has been discussed in one of the many subthreads, but is there a reasonable explanation for the finding that the crashed Tesla "deleted its local copy" of the crash data?

> Within ~3 minutes of the crash, the Model S packaged sensor video, CAN‑bus, EDR, and other streams into a single “snapshot_collision_airbag-deployment.tar” file and pushed it to Tesla’s server, then deleted its local copy.

Putting aside the legal implications wrt evidence, etc — what is the ostensible justification for this functionality? To a layperson, it's as bizarre as designing a plane's black box to go ahead and delete data if it somehow makes a successful upload to the FAA cloud server. Why add complexity that reduces redundancy in this case?

Two that come to mind:

1) Embedded systems typically do not allow data to grow without bound. If they were going to keep debugging data, they'd have to limit it to the last N instances or so. In this case N=0. It seems like the goal here was to send troubleshooting data, not keep it around.

2) Persisting the data may expose the driver to additional risks. Beyond the immediate risks, someone could grab the module from the junkyard and extract the data. I can appreciate devices that take steps to prevent sensitive data from falling into the hand of third parties.

It would be trivial to set it up to only delete old instances when free space goes below a threshold.

If the data can expose the driver to additional risks, then the driver can be exposed by someone stealing the vehicle and harvesting that data. Again, that can be trivially protected against using encryption which would also protect in the instance that communication was disrupted so that the tar isn't uploaded/deleted.

One of the few reasons I have _never_ entertained the possibility of buying a Tesla: despite collecting enough data to know what happened in almost every crash, the company will withhold the data if Tesla is at fault, but is more than happy to quickly publish selective data to support a PR assassination of a driver who is at fault.

There are obviously other reasons, but I would much rather buy a car from a company that either doesn’t collect my data or doesn’t feel the need to bury me in the media if I screw up.

I would like to see how the decision was justified to implement automated deletion of the onboard snapshot.

https://en.wikipedia.org/wiki/Tampering_with_evidence

.. a person ..

There aren't enough details in the somewhat hyperbolic narrative format to really say, but if I were going to create a temporary archive of files on an embedded system for diagnostic upload, I would also delete it, because that's the nature of temporary files and nobody likes ENOSPACE. If their system had deleted the inputs of the archive that would seem nefarious, but this doesn't, at first scan.

"Their system" is a car, sold as a consumer product, which has just experienced a collision removing it indefinitely from normal operation. Reconsider your analysis.

Yes? But the article doesn't say that Tesla deleted the EDR, it says they uploaded the EDR file in an archive format, then deleted the uploaded entity. Which strikes me as totally normal.

Totally normal for a completely different domain. Very abnormal for what's functionally the car's black box.

No, the car's "black box" is the EDR, the behavior of which is regulated by federal agencies. This article is discussing ephemeral telemetry which accessed the EDR.

No, the EDR forms part of the car's "black box" – just like the FDR forms part of an aeroplane's black box. Per the article, the erased* telemetry ("collision snapshot") contained quite a bit more data than just from the EDR.

*: I can't work out from the article whether this file was erased, or just unlinked from the filesystem: they quote someone as saying the latter, but it looks like it was actually the former.

So? Tesla needs to bear the responsibilities of having deleted its own potentially exculpatory evidence. Not granted an inference that it did so and therefore Tesla is innocent.

Tesla was having issues with log file writes destroying their chips. They can argue they have precedent for deleting data, but not hiding it.

Wouldn't creating an archive on the filesystem and then deleting the archive cause more writes than just creating it without a delete?

I think the goal is to save as much as you can in the interim. Holding onto X bytes of archives is more time worth of data than X bytes of uncompressed. We do that stuff all the time in finance. Stuff gets spewed off to external places and local copies get kept but archived and we simply rotate the oldest stuff out as we go. If the cleanup process is configured separately from the archiving process you can absolutely archive things just to remove them shortly thereafter.

Wear increases when free space is low as there's less overall space to put the data. If you only have 500MB of free space, those blocks take the majority of write hammering until the chip fails. If there's 5000MB free, you can spread the wear.

Not in the big picture.

Fundamentally, flash memory is a bunch of pages. Each page can be read an infinite number of times but there are quite relevant limits on how many times you can write it.

In the simplistic system lets say you have 1000 pages, 999 hold static data and the last one keeps getting a temporary file that is then erased. All wear occurs on page 1000 and it doesn't last very long.

In the better system it notes that page 1000 is accumulating a lot of writes and picks whatever page has the least writes, copies the data from that page to page 1000 and now uses the new page for all those writes. Repeat until everything's worn down. Note the extra write incurred copying the page over.

In the real world a drive with more space on it is less likely to have to resort to copying pages.

I would be fascinated to entertain arguments for how the future write life of a flash memory chip, meant for storing drive-time telemetry in a wrecked car, merits care for preservation.

Many parts, especially non-moving parts, hold value after a wreck. It does not make sense to keep the file on the disk for the lifetime of the part. The part is most likely going to be resold as used. The part will have a shorter life if the big crash data file is left on disk.

I didn't know Tesla authorized any resale of parts from any of their vehicles, used, crashed, or otherwise! Certainly I had no idea they built support for it into the system software. Was that a recent change?

https://Service.tesla.com has been online for I think 2 years

Oh, I'm sure it has. Owners tell a different story about what the site is worth. Of course, Apple actually "innovated" that one. But do you seriously mean to suggest Tesla engineer their cars to discard potentially dispositive crash analysis, in order to support their own resale of junkyard pulls?

What's the alternative? Keep it on there for the next owner to view?

Okay, let's recap.

Your implicit claim is that there exists a resale market for wrecked Teslas, meriting such product and engineering interest from the company that the car's data storage, in the immediate wake of a disabling collision, preemptively and correctly destroys information of use to crash investigators, so that the company can cannibalize the wrecked hulk for items which Tesla then goes on to sell as new OEM repair parts.

Isn't it embarrassing to go to all this? It certainly seems like it should feel that way, for as degrading as it looks from the outside. If you can explain it, I would like to understand what makes it seem worth your while.

The data was not preemptively destroyed. It was uploaded and then deleted. The implications for deleting are obvious. The part can be sold and reused after the crash data is erased. The part should not be sold and reused until that information is erased because it contains a large file and this file will likely contain death footage.

> The data was not preemptively destroyed. It was uploaded and then deleted. The implications for deleting are obvious. The part can be sold and reused after the crash data is erased. The part should not be sold and reused until that information is erased because it contains a large file and this file will likely contain death footage.

Okay, then your contention appears that Tesla is content with not bothering at all to refurbish the junkyard pulls it sells through OEM channels as OEM repair parts, thus denying themselves any opportunity to manually wipe sensitive data - potentially to include video of the prior owner's violent death! - as part of any remanufacturing or even basic QC process. But - as has now been made a matter of public record, in consequence of their fighting and losing this wrongful death case - Tesla do make sure to keep a copy for themselves, one of an apparently large collection of same which they went far out of their way for years to keep anyone else from discovering even exists.

Is there anything you care to add to that? Feel free to take your time. You've said quite a lot already.

This counterargument does not hold. The vast majority of crashed vehicles are not sent back and resold by Tesla. They are resold at auto auctions, junkyards, ebay, etc. Deleting the crash data protects Tesla from lawsuits on both ends of the chain: the victims and the new owners, while also preserving the part's lifetime.

> Deleting the crash data protects Tesla from lawsuits...

Oh, obviously.

When a vehicle crash occurs, that embedded system should no longer be treating data as "temporary", but as what it now is, civil and potentially criminal evidence, and it should be preserved. To go to the effort of creating that data, uploading it to a corporate server, and then having programming that explicitly deletes that data from the source (the embedded system), certainly reads as nefarious without easily verifiable evidence to the contrary. The actions of a company that has acted this way in no fashion lends any credibility to being treated as anything other than a hostile party in court. Any investigators in the future involving a company with such a history need to act swiftly and with the immediate and heavy hand of the court behind them if they expect any degree of success.

Maybe, if this were a toaster or fridge.

The main reasons to store data are for safety and legal purposes first, diagnostics second. Collision data are all three. They need to be prioritized above virtually everything else on the system and if your vehicle has had so many collisions that the filesystem is filled up, that's a justifiable reason to have a service visit to delete the old ones.

If I were implementing such a system (and I have), I could see myself deleting the temporary without much thought. I would still have built a way to recreate the contents of the tarball after the fact (it's been a requirement from legal every time I've scoped such a system). Tesla not only failed to do that, but avoided disclosing that any such file was transferred in the first place so that the plaintiffs wouldn't know to request it.

The tar file is a convenient transport mechanism for files that presumably exist in original form elsewhere within the system. (All bets off if sources for the tar were changed afterward.)

Given storage is a finite resource, removing the tar after it was confirmed in the bucket is pure waste.

I'm not sure whether you're saying that the tar should or shouldn't be deleted. Regardless, my point isn't that it was intentionally deleted. I can easily imagine someone writing a function to upload the data using something like std::tmpfile (which silently deletes the file when it's closed) without thinking about the full implications of that for the broader context the code exists in.

Even in that case though, you would still have a way to produce the data because it would have been specced in the requirements when you were thinking about the broader organizational context.

For enospace, I’d put telemetry data on an entirely different partition or even device, as to isolate the rest of the filesystem.

What about not handing the tar ball to the police looking for data. Do you see that as a problem?

That's obviously problematic. I am only commenting on the belief in a conspiracy of programmers here. The overwhelmingly most likely reason that a temporary file would be unlinked after use is that is what any experienced systems programmer always does as a matter of course.

What "belief in a conspiracy of programmers?"

I've made no contention, but if I had, it would be that whoever signed off on this design had better not have a PE license that they would like to keep, and we as an industry would be wise not to keep counting on our grandfather-clause "fun harmless dorks" cultural exemption now that we manufacture machines which obviously kill people. If that by you is conspiracy theory, you're welcome.

There are not PEs signing software changes for ancillary equipment in consumer vehicles.

ETA: Restate your conspiracy theory in the hypothetical case that they had used `tar | curl` instead of the intermediate archive file. Does it still seem problematic?

"Ancillary" is quite a term for crash reporting after a crash. That is, for a system responding as designed to a collision which disabled the vehicle.

I'm not going to argue with someone who throws gratuitous insults. Rejoin me outside the gutter and we'll continue, if you like. But the answer to your question is trivially yes, that is as professionally indictable, as might by now have been clarified had you sought conversation rather than - well, I suppose, rather than whatever this slander by you was meant to be. One hopes we'll see no more of it.

It seems they waited for a subpoena. Would you prefer automakers send the police a notification anytime the car records a traffic infraction, or maybe they should just set up direct billing for municipalities?

I would prefer automakers to not pretend that they don't have telemetry that they actually do.

It isn't clear that Tesla pretended anything aside from a claim in the article. The article: 1) does not directly cite an original source; 2) walks back the claim to "acted as if". The police had access to everything in vehicle the entire time. The civil lawsuit plaintiffs also had access to everything as evidenced by their forensic expert's access. That Tesla also received a copy of the vehicle's data is not relevant.

Don't Talk to the Police: https://www.youtube.com/watch?v=d-7o9xYp7eE

The police reached out to Tesla legal within days and asked if they needed a subpoena. The Tesla lawyer said no, but then directed the police to make their request in a specific way that hid relevant data from police and the victims. It took 5 years to get the data, and only then after a forensic engineer proved Tesla uploaded it by finding logs on the cars black box and after the court was leveling sanctions.

Should companies be intentionally misleading to law enforcement about what data they have on fatal accidents like Tesla?

I hadn’t listened to the below video in a while but I did again today and the above comment’s assumptions of fact perfectly illustrate the point of the video.

This comment takes as fact a claim made by the police, which might be wrong, either by error or purpose.

One thing is for certain though, the police’s initial investigation was criminal. That is to say it was to establish the fact of who was the driver, etc. It was totally separate from the civil litigation that later established Tesla to be at 30% fault for the wreck using computer records establishing that ADAS was engaged.

Leaving Tesla out of it, suppose there was some other automaker with some other problem. A cop comes to them and says “what do I need to ask you to establish the facts about an accident?” Since when do police just accept “oh, the potentially adversarial lawyer gave me advice on what they can tell me without getting a subpoena. Guess that’s all I can do?” That’s absurd.

Don’t talk to the police: https://youtu.be/d-7o9xYp7eE

I would love to see what you need so much disk space for after the car is crashed and airbags are deployed. If that event fires the car is going in to the shop to have its airbags replaced at a minimum. Adding a service step to clear up /tmp after a crash is fairly straitforward.

You're assuming that it's a separate routine. It could very well be simply a routine that uploads any incident.

And deleting the file would have been perfectly fine, if Tesla had given out the file when asked for it. Instead of what they did.

What was deleted was the compiled file sent to Tesla, not all the bits of data that it came from. Nothing malicious in that, just code not leaving trash around.

Well, thank goodness for that. Wouldn't want to die in a car crash with an untidy filesystem, just like Mama always said.

This is baffling. Those files are the most important information the car contains after an accident.

Let's not be naive, or deceptive, about a malpractice from a multi billion company owned by a multi billionaire.

They did not erase the original data, only the compiled file that was sent. That's just proper coding, clean up your temporary files!

The file was reportedly named "snapshot_collision_airbag-deployment.tar", which indicates that the developers are able to know that this particular event is a terminal one. Is it really necessary to be mindful of a few GB of storage space on a device that will likely never again to record data for its host vehicle?

Whatever named the file knew it was a bad event, but that doesn't mean the routine to transmit an incident had a special case for this.

What do you call two lying executives in minimum security prison? A good start.

So the solution to all of this is to lock up more executives who commit fraud or lie to the public.

Recently (after a 10 year battle) two former Volkswagen executives just got prison time for the Dieselgate scandal.

Wish it had come faster, but that's a good start.

https://www.lbc.co.uk/crime/volkswagen-execs-jailed-fraud-de...

It's difficult for me to tell in the article because how much the terms are used interchangeably, but it was it FSD or autosteer that was driving the car when it crashed?

My autosteer will gladly drive through red lights, stop signs, etc.

And the fact that we have telemetry at all is pretty amazing. Most car crashes there's zero telemetry. Tesla is the exception, even though they did the wrong thing here.

Autopilot, not autosteer. Wording here is important.

This was in 2019 so I don't think FSD was a thing yet.

"...the investigator thought that Tesla was being collaborative with the investigation at the time..."

So this is also a failure of the investigator.

No its not. Investigators will try to obtain data willingly unless they are made to believe that the party is not acting in good faith.

I suppose it isn't but the sheer scale of effort seems like it should be criminal.

Lying to the police is illegal, which it seems like Tesla (employees) did many times.

It is absolutely criminal. Whether it gets prosecuted is a different matter - prosecutors love giving corporate interests a free pass

6 months ago it had no chance of being prosecuted. But, if the fallout between Elon and Trump is as bad as it looks from the outside, there might be justice after all.

I doubt Trump cares about Musk's flunkies.

The longer the farce goes on, the more I think the laggers in the self driving car industry are more trying to wait out regulators than actually get good enough.

That is - gamble GOP alignment leading to to regulatory capture such that the bar is lowered enough that they can declare the cars safe.

Despite what you hear from certain media voices, there are effectively no performance-based regulatory barriers in the US. You can claim any autonomy level you want at any time and aside from the small number of states that have a real permit process that rises above a rubber stamp (literally just California), regulators are reactive to headlines of your system failing, not its actual performance.

Even California's system is lax enough that you can drive a Tesla semi through it.

It's really good that these stories start to hit the headlines - Tesla was doint it for ages. Only recently they start to feel the heat and settled some cases. This should criminal behaviour.

Another interesting thing is that this post does not appear on HN front page. Surely it should be relevant here that one of the top AI / Robotics / Tech, whatever you name it, company in the world is doing this kind of stuff in the open. But that may get people curious to actually look at what and how they are doing things and realize it's all smokes and mirrors and maybe there is much more shady business going on there.

I just can't wait when this house of cards collapse.

https://www.courtlistener.com/docket/59932667/benavides-v-te...

> Tesla fans need to do a quick exercise in empathy right now. The way they are discussing this case, such as claiming the plaintiffs are just looking for a payout, is truly appalling.

> You should put yourself in the family’s shoes. If your daughter died in a car crash, you’d want to know exactly what happened, identify all contributing factors, and try to eliminate them to give some meaning to this tragic loss and prevent it from happening to someone else.

> It’s an entirely normal human reaction. And to make this happen in the US, you must go through the courts.

This (especially the very last point) is crucial. Whenever there is any kind of error or mistake by a big corporation, more often than not, its immediately covered up, nothing is admitted publicly. But when a lawsuit is involved, the discovery process will lead to the facts being uncovered, including what the company knows.

I am glad that they were able to uncover this. Someone I know lived in an apartment complex that was made uninhabitable due to an obvious fault on the owner, but they didn't get a straight answer about what happened until they sued the owner and got the details in discovery, a couple of years after the incident. This is the only way to get to the facts.

I am rather sick of the AI generated hero image, but this one made me laugh.

It was probably made using Musk's own Grok as other AIs disallow depicting real people.

I think it's wild that a legit article would use an image like that. Sure it's funny, but save it for social media, not a news source that's supposed to be based on facts.

This is why Waymo will win. They've focused on transparency and building trust. They understand they are operating in the most physically dangerous space most people ever encounter in their day-to-day lives. That, and their technology actually works.

Tesla is comparatively a bull in a china shop. Raise your hand if you would trust Tesla over Waymo to autonomously drive your young children for 1,000 miles around a busy metro. That's what I thought.

normally a person would go to jail. But corporations just pay a fine. I think we really need to come up with punishments that are actual deterrents. Like any time a corporation ends up killing someone from negligence there needs to be an action that is equivalent and scaled appropriately.

Send the corporation to jail. That means it cannot conduct business for the same amount of time that we would put a person in jail.

If "corporations are people" then there should be a way to incarcerate and/or execute criminal corporations. And we the people should do it roughly as regularly as we incarcerate/execute actual human criminals.

Corporations don't actually exist, except on paper. Preserve the records, and you can un-execute them if need be. This suggests we should be executing corporations more readily than we execute humans.

Possibly, but start with prosecuting humans (company staff and directors) as accessories to the crime.

No need for anything so drastic, just make the fines a sufficiently large percentage of corporate free cash flow.

Make negligence unprofitable and the profit-optimizers will take care of the rest. Then 80 years later when people get too used to "trustworthy" corporationns we can deregulate everything and repeat the cycle

The last time that a major corporation was found guilty for criminal behavior (Arthur Andersen in 2002 for Enron related stuff) the company closed immediately. This has led to problems in the audit industry, where was once the Big 8 audit firms has shrunk, due to mergers and AA dissolving, there basically are barely enough firms to independently audit each others books, and it's made the audit market much worse.

The MCI Worldcom fraud, which broke shortly after Enron, might also have doomed AA (they were the auditor for both major frauds of 2002). MCI Worldcom filed for bankruptcy before it could be hit with criminal charges, and the SEC ended up operating MCI-W in the bankruptcy, because the fines were so large and are senior to all other debts, so they outmuscled all of the other creditors in the bankruptcy filings. Which was why they weren't hit with criminal charges- they already belonged to the Government. There hasn't been much stomach for criminal charges against a corporation ever since.

The fact that the Supreme Court has spent the past few decades making white collar crimes much harder to prosecute (including with Arthur Andersen, where they unanimously reversed the conviction in 2005) is another major factor. The Supreme Court has always been terrible, and gets far more respect than it deserves.

Would be good for Tesla's reputation if they would upload/mirror the crash data to a neutral third party. Then the police and plaintiff could get it from there without any allegations of withholding data.

Perhaps it's time for Tesla to face Judicial Dissolution[1], aka the Corporate Death Penalty. It would send a clear message to Musk, and those like him responsible for these type of unacceptable acts.

[1] https://en.wikipedia.org/wiki/Judicial_dissolution

If this 3-year-lie parade is what Tesla Inc does for killing ONE PERSON, imagine the lengths it will go to concealing MASS DEATHS it is in any part responsible.

Tesla deserves regulating.

anyone with a Tesla using this or something similar these days? curious if it would have been helpful in this case.

https://github.com/marcone/teslausb

That's highly illegal. I don't understand why no one was criminally charged.

There are a painful number of typos in this article; whose content is otherwise good.

Of course they did. This is how Tesla has been operating for a very long time.

I get this may be off topic, but does anyone think these cheesy, bad AI-generated headline images help the article's point of view, or heck even make it more engaging?

It just looks stupid to me in a way that makes me more likely to discount your post.

It started with YouTube thumbnails and leaked from there. It just gets more clicks compared to a legitimate photo

I can get it for some rando YouTube video.

For an article that is supposed to at least smell like journalism, it looks so trashy.

Journalism? It's literally just a blog—a very successful car-influencer blog that's in the past earned 6-figure payments from Tesla itself[0], for their very successful shilling of Teslas.

Journalism is a thing of its own; blogs aren't it.

[0] https://www.thedrive.com/news/24025/electreks-editor-in-chie... ("Electrek’s Editor-in-Chief, Publisher Both Scoring $250,000 Tesla Roadsters for Free by Gaming Referral Program": "What happens to objective coverage when a free six-figure car is at play? Nothing good." (2018))

I haven't done it, so I don't have any data to back it up. I suppose it works, at least short term, that is why so many websites, video creators, copywriters, email newsletter writers etc use it?

Negative, cheesy, clickbait, rage inducing etc headlines do seem to get more clicks. There is a reason why politicians spend more time trash talking opponents than talking positively about themselves. Same goes with attack ads.

I’m with you there. Depending on other factors on the page, it can be a smell or a red flag.

OpenGraph requires an image for your article regardless of if it is useful or not, for the sake of embeds in facebook and discord (and others)

There's nothing requiring that image to be unique. Lots of sites just provide a higher solution favicon.

People like to think they love literature, yet they read tabloids. People like to say they want better information, yet get their news from social media.

I have not doubt a majority of people will say they despise these pictures like YouTube thumbnails, yet the cold numbers tell the opposite.

This one is a such a bad photoshop too! The box's text is clearly AI generated, with an older model, and the "Autopilot crash data" is imposed on it with an image editing tool. Really cheap looking.

If you are not familiar with electreck this is basically a nutshell of their business. Tesla bad = clicks. Tesla good = no clicks. It is stupid and it works, as you can clearly see in this particular example in YC

Jail the CEO.

How strange.

"The secret of a great success for which you are at a loss to account is a crime that has never been found out, because it was properly executed."

       - Honoré de Balzac

"Tesla awards boss Elon Musk $29bn in shares" - https://www.bbc.com/news/articles/cz71vn1v3n4o

given these: https://www.youtube.com/watch?v=TMk28nV9nkI and how connected Elon is to Trump ... I doubt anything will come of it.

After their recent falling out though it might be the other way around no?

This article is about a civil suit where they got fined $300M dollars

How do you go and correct someone to say that it was a civil suit and then call the judgment a fine???

Funny I was going to fix that earlier but figured it was close enough to provide OP context. Being 100% right on the internet is not that big of a deal. Although it's true smugly nitpicking text in people's comments is not helpful behaviour.

I think it's a pretty meaningful distinction but perhaps I'm in the minority there. I certainly don't care to be 100% right on the internet about everything.

buried under all that sensationalism is this line... "The driver was responsible for the crash and he admitted as such..."

I’m not knowledgeable of this incident, so let’s assume I accept your argument that it was the drivers fault (seems likely enough).

Are you also arguing that Tesla didn’t withhold data, lie, and misdirect the police in this case, as the article claims? Seems to me that Tesla tried to look as guilty as possible here.

The driver was found 2/3rds at fault, Tesla 1/3rd.

I agree with you that doesn’t matter when it comes to covering up/lying about evidence.

They could have been 0.5% at fault. Doesn’t mean that was ok.

Bullish for TSLA.

Cruise had to shut down after less than this but, because Elon has political power over regulation now, a Tesla could drive right through a farmers market and they wouldn't have to pause operations even for an afternoon.

> Elon has political power over regulation now

Does he still? I wouldn't be so sure.

I probably wouldn't care much about Tesla and Elon, but he scammed his way to become a fortune 500 company which made SPY and my 401k autobuy Tesla.

Buying SPY, my mistake. Being incentivized to put money in my 401k... That is a bit harder to solve.

It's actually quite easy to solve, you can calculate how many shares of tesla you own via the index fund and buy an an equal quantity of TSLS.

'Easy'

Well, I didn't know about this. I didn't consider it. I also would then need to spend a non-tax deductible amount on this. And IIRC Tesla entered the Fortune 500 before 2022 when that TSLS started. Then I'd have to actually do the calculation, and continue doing the calculation as SPY/my 401k adjusted each month or so.

'Easy'

Reply on news.ycombinator.com