The thing missing is somebody to define that severe security vulnerability is considered as a defect in the product. Once this is done, consumers can either demand the seller to fix the problem, replace product or give refund. This should pretty quickly create a financial incentive to sell products which get updates.
Still, yea, this is exactly the kind of thing that needs to become commonplace. Hopefully it will be, though this is far from the first time that something like this has happened and here we still are.
Easy to do the same thing here, as long as the company has (and wants to keep having) presence in the EU that's enough to enforce standards worldwide.
Still, even just the fact that companies are now thinking about this and providing ways to download/delete your data makes me love the GDPR.
It would be just the question of somebody testing it. Take a relatively new() device with unpatched security hole, try to get it replaced/refunded and if necessary proceed to court to get a decision on the matter.
Although I wonder if any vendor actually wants to go to court arguing that a product with severe security issue is actually working as intended.
() Safe bet should be a device purchased less than 2 years ago, since that seems to be the time the vendor is at least responsible for defects.
Or an incentive not to sell in Europe.
Of course that doesn't fix the global problem.
Let's hope this forces manufacturers to improve their security. That's the main good that needs to come out of this.
They have generally had mixed success, from the ones I've read about. Pretty often there are side effects that aren't good (e.g. bricking some devices, excess traffic due to being too successful, that kind of thing)
But yea, in this case I think it's clearly too complex. There are thousands of different things hooked up to the internet, and you can't fix all of them at once.
If they wait too long, then the result is more sales, which increases profit for the whole ecosystem that produces these devices. But if it's more common and results in returns and exchanges, the result is less profit, which might finally put some pressure where it needs to be.
You'd need a JTAG cable and a fresh firmware image to fix that. Hardly something within the reach of a typical home user.
It's an interesting problem in that the value of hackale/wormable IoT devices stems from their sheer number. In a way the culpability of any singular vendor for any singular sale is low by and of itself; it's only large in aggregate of all the vendors' products ever sold & still deployed.
This is a reverse of, but comparable to, the low value of personal data being gathered&processed (or stolen) - for every singular individual, the value/loss is exceedingly low. The value lies in the aggregate of the data; the whole is much more than the sum of its parts. Thus the prosecution of things like the Equifax hack, or the insecure IoT devices is pretty spotty at best.
Perhaps we need a new, specialized legislation & judiciary for cases where network effects dominate.
That said, I agree there's a sort-of similar pattern here, which I feel underlies many of the biggest problems of our era - including climate change. The pattern is that there's a lot of entities - individual, small and large companies - engaging in transactions, and each transaction has a small negative - some personal data stolen here, some toaster joining the botnet there, some trivial amount of carbon emitted elsewhere. The negatives however add up, and ultimately manifest as huge and international problems.
We definitely need to develop legislative methods that would combat this pattern at the structural level, regardless of the business domain it shows itself in.
That doesn't entirely match the current situation, but there probably is a some responsibility that lays with consumers for buying crappy devices from companies with no track record for keeping them updated. It's not always simple to reason about, but it is something to consider next time there's a decision to buy some cheaper device from some no name company or from some well known company (that hopefully has some sort of support lifetime at least).
When appropriate measures have massively failed, what are the alternatives? Vigilante actions are one option, as is "watch the whole thing burn down", and probably an infinite variety of stuff in the middle and along different axes. What's your preferred option?
edit: also, it's not quite "may [do] something worse". Insecure IoT things are used as botnets. Frequently. It's not a hypothetical threat at all, it's just a question of scale / frequency.
So, vigilantees concerned about damage to innocent people keep breaking into the windows, stealing the guns, unloading them, and tossing them into landfills. I'd be like: "Stop putting your guns in front of the windows. Be a responsible gun owner." Enough broken windows and stolen guns might incentivize them to do that.
So, in a way, yes. I do. So do lots of people when they go to urgent care rather than the ER, knowing that the ER could bankrupt them, and they'd rather risk the delay. I don't have numbers off the top of my head, but I don't think it's as uncommon as you seem to think it is.
"People", in general, are dumb. Individuals, when explained, might actually support it, but I guess you'd have to start teaching the following in school:
If you want to compare things, you need to express them with the same unit. It so happens that there is universal unit already in use to compare everything to everything else - money. Which, internationally, means dollars. Moral caveats apply, but this is how you correctly compare arbitrary things when you lack a more suitable common unit.
> Like, seriously? If you went around and asked people whether their restaurants should be shut down for unsanitary practices, do you think you would get a response on the same planet as what you would get if you asked them about their IoT devices getting shut down? Come on.
Sure, why not.
Unsanitary practices: you weigh some people likely ending up in the hospital with food poisoning against shutting down a business employing a dozen or more people for unspecified amount of time, which may or may not have pretty bad secondary effects (e.g. waiters aren't exactly the kind of people who can afford asudden job loss).
IoT botnet: you weigh bricking a bunch of (at this point in time) non-essential trinkets, thus inconveniencing innocent consumers and making it worse for garbage companies who shouldn't be in business in the first place, vs. enabling moderate-probability inconveniencing events for thousands to millions of people, such as DDoSing random websites, and occasional low-probability high-impact event, such as Maersk getting pwnd and disrupting worldwide shipping.
Ultimately I'd expect the IoT scenario to be less important than food poisoning scenario when you tally up both sides of both choices - and accordingly, food poisoning is something governments worldwide are already dealing with. However, and circling back to tty2300's original comment, the pattern of thinking is pretty much the same here in both scenarios.
No. I'm responding to:
>You measure your health in dollars?
And apparently reading it with a different intent than it was written with.
---
>If you went around and asked people whether their restaurants should be shut down for unsanitary practices, do you think you would get a response on the same planet as what you would get if you asked them about their IoT devices getting shut down?
They should, yes. I don't really think they do though. Botnets cause them harm (they're often used to hide the source of hacks, which often target financial info, which does affect millions of people), but it's often further removed than food poisoning, so it's harder for them to identify the cause. They're also usually more attached to purchases of things that work for longer than a meal lasts.
People don't have visceral reactions to things in line with what they should rationally fear. I don't think anyone would debate that. I'm arguing DDoS (and everything else botnets are used for) are in that "does not get an appropriately strong reaction" category.
You are priced per body part based on your income. Your health is most certainly defined in dollars.
I daresay that this situation even applies to "most people", or at least a very sizeable proportion of the global workforce.
It's IMO somewhat comparable to culling high-risk farm animals to prevent widespread disease outbreaks.
> Where do you have legal authority to do a vigilante culling of your neighbor’s herd, because you decided something about them merited it?
I am not making the argument that this is legally defensible. You can by definition not have legal authority to engage in vigilantism. It's from an ethical and practical perspective that I draw the analogy.
On the matter of legality, the difference you point out applies, but also goes both ways. There is no legal framework around "culling" insecure devices, and very little regulation.
Where do you have legal authority to do a vigilante culling of your neighbor’s herd, because you decided something about them merited it?
>The BrickerBot author argued that it would be better if the devices were destroyed, rather than sit around as cannon fodder for DDoS botnets, and haunting the internet for years.
... yea, broadly I'd agree. IoT vendors are causing a tragedy of the commons, inflicting quite a lot of damage without feeling any of the pain because it hits others.
It's the sort of thing that should be addressed by legislation of some kind, but absent that (which includes nearly all international cases)... what else can you do to stop the worst offenders?