Practically all modern vehicles are privacy nightmares on wheels. They have identifying sensors all over the vehicle, on both the exterior and interior. And all ECUs (monitoring and controlling sensors and actuators from a vehicle’s particular subsystems) are (indirectly) connected to a cellular modem, which is likely to support ‘vehicle to everything (V2X)’ communications: where the vehicle exchanges sensory data with external networks (primarily the manufacturer’s), road infrastructure and other road users nearby.

Your most effective remedy is to hamper the functioning of the modem, that is if the manufacturer doesn’t throw an error code as a result of it. If that isn’t an option, check if it allows you to physically disable sensors (taping off, or limiting the FOV of cameras, or unplugging microphones). And lastly, if even that isn’t an option, disable as many “features” as you can; even though some may require you to manually re-disable them, every time you restart your vehicle…

Keep in mind that in doing so, you might also disable actual features (ones you might find genuinely useful). Mozilla’s somewhat outdated reports of manufacturer’s privacy statements might give you some idea, of which brands could be considered the lesser of evil. But these are subject to constant change, as are the over-the-air updates of vehicles in their fleet, and their dynamic user manuals…

All the time. There’s occasionally even websites that outright block me without challenge, run many iterations on my machine, or present a tiresome amount of CAPTCHAs (Waymo machine vision must be trained pretty well, from me labeling goddamn traffic lights, motorcycles and stairs throughout the day). Especially CAPTCHAs that load new pictures with an infuriatingly slow fade animation drive me up the wall; one of which possibly not even sufficient.

But I take it as a compliment, as it might mean my browser profile is minimal enough to trip these systems into thinking it’s a headless browser (typically used for bot traffic).

Authentication of users seems to be outsourced to ‘identity providers (IdPs)’ (primarily: Google, Facebook, Apple, etc.: “login with…”); that manage access to a service’s resources, which are increasingly distributed throughout various cloud-based services. A service’s chat application might be hosted at provider X, whereas its website at provider Y; so managing access to these resources from a centralized identity pool is convenient. Well, unless you value you privacy, but that’s the end-user’s problem…

At least that is the impression I’ve gotten from my limited research on Amazon Cognito: required for contacting my doctor’s office through their forced chat application (automatic account creation, no longer a phone line for uses outside of emergencies). And yes, to anyone wondering: I am going to look for alternatives, because this should not be acceptable, at least for a doctor’s office of all places.

“Consent” from just the wearer, that is likely to have their iPhone’s photos/videos synched to their iCloud account: which by default don’t appear to be end-to-end encrypted (and is therefore accessible by Apple). So although the user is technically expected to read Apple’s entire privacy policy to everyone they interact with, and they technically can enable E2EE on iCloud for this category of data, they know full and well the ordinary iSheep isn’t going to.

Given Apple’s obsession with brand-perception, risking being associated with perverts (like rightfully happened with Meta’s “pervert glasses”), I think it’s completely inexcusable to release a product of the same category; at least for the time being. But perhaps Apple’s marketed propaganda has really given the ignorant populace a false sense of privacy, in which they blindly trust a Trillion dollar corporation with their most privacy-sensitive information.

An application that would make donating to FOSS projects, especially GPL-licensed works (or at least: those not trying to cater to corporations) in desperate need of support, effortless. In that: a periodic budget can be determined, time spent actively using applications is tracked, an application’s userbase is taken into consideration (as their cumulative income may eat too much into budgets: possibly better spent supporting smaller projects), and personal preference can be given (including an option to filter license types); as to rank the applications based on perceived value, and fraction the budget accordingly.

I would be happy to support projects that I actively, or to a lesser degree, rely on. But the process of taking all steps required to support a project is simply too manual, and outright unmanageable, when trying to take into account all system components. I really want the incentives to shift from MIT to GPL, so that maintainers don’t effectively end up being controlled by corporate paychecks; and are rather supported by individual end users. This should in theory make FOSS more resilient to any future restrictions to source access, which could hamper the long-term parity of forks, seeking to strip anti-features.

Here’s the link you may have forgotten to add: https://www.marketwatch.com/story/employers-are-using-your-personal-data-to-figure-out-the-lowest-salary-youll-accept-c2b968fb

If a local ‘date of birth’ field defaults to ‘18 and up’ if unset; the bracket is communicated using an open, standardized format; specification is on a trust-basis, without external verification required; and the field automatically unsets upon turning 18; I might not be that opposed to an implementation like this.

But without a requirement for legal presence within the state of Illinois, any application developers without it, cannot really be held liable for any infringements on it. Bad actors might also learn which internet users are of which age bracket, and this all might be the beginning of another very slippery slope, where requirements keep building as a result of misuse and/or ineffectiveness.

Okay

You understand this is a perverse foot in the door, right? When people become accustomed to the idea, of their every conversation possibly being recorded (for whatever reason), the follow-up step is bound to be cloud connectivity: with limited on-device storage and lack of redundancy (back-ups), cloud will naturally serve as the “convenient” solution for both of these self-created “problems”. Problem-reaction-solution, always.

DuckDuckGo Search simply guesses your location using a GEO::IP lookup with the IP address that’s automatically sent to us via your device; then we throw away the IP address. […] The next time you perform a local search, selecting “Use Location” will send us a random location nearby, while hiding your actual location from DuckDuckGo. […] If you’d rather choose a custom location instead, you can pick “Set Location on Map” instead of “Use Location.” This bypasses the browser location permission, but still keeps your precise location anonymous to us.

I’ve just switched back to Brave, because (No AI) DDG keeps localizing results, despite region filter set to “All regions”; which really bugs me. If you want to disable Brave’s annoying AI “feature” (the reason I switched to DDG), without having to store a cookie: install the Redirector addon and ‘Create new redirect’ with:

Example URL: https://search.brave.com/search?q=test

Include pattern: https://search.brave.com/search?q=*

Redirect to: https://search.brave.com/search?q$1&summary=0

Sounds noble, but I have a gut-feeling it will simply result in more expensive electronics; which simply aren’t expensive enough as is… If an OEM is required to provide parts for 10 years for each device it sells, you better expect them to pass that additional cost onto the consumer. Besides that, there are already manufacturers providing parts for devices, but the high relative cost of parts make repairs unattractive, compared to simply “upgrading”. And the cheap segment of devices might disappear altogether, even more adware comes preinstalled onto them, or this line of devices is being subsidized by higher segments. Or perhaps more OEMs will lean into consumers leasing devices on a subscription basis, which would also compensate for the loss in sales for new devices (if actually effective); and seems in line with the anti-ownership trajectory. :)

Great article. It touches on the primary concerns I have with modern surveillance tech, namely:

This “privacy for me, surveillance for thee” attitude feels like a perfect encapsulation of the future we’re already living in: one where wealthy elites can afford privacy, while the rest of us are forced to live in a corporate surveillance panopticon.

This concept has crossed my mind as well: privacy becoming a privilege reserved for the super-wealthy; it simply being added to luxuries the underclass should naively strive for, in exchange for their exhaustible time and energy.

AI surveillance in the hands of wealthy individuals could create opportunities for Epstein-like abusers to get away with their crimes, allowing monied interests to get dirt on politicians, law enforcement and even judges.

Wealthy individuals getting away with crimes is the most terrifying to me. Where either knowing mazes in surveillance infrastructure, or systems explicitly turning a blind eye, could create situations where: there’s no evidence placing the actual perpetrator at a crime-scene, while an innocent person, simply being in the wrong place at the wrong time, is framed with evidence to support it.

Thanks for reminding me to go into settings and uncheck ‘Show Bot Accounts’. Why is this a feature anyways?

Their magazine on the subject is quite fascinating. Especially the ALPRs disguised as traffic cones (page 33), and trail cameras in trees, surroundings, trash, and even signs and posts (page 35 and 36) are as interesting as they are creepy…

I belief it’s morally wrong to do that, both from the perspective of the donor and party: as people democratically voted to land the party the low number of seats it has. If a donor could just dictate party policy, you’re disregarding the promise made to the voter. For such an insignificant party it’s likely to give them additional budget for campaigning, so they might have a better chance at competing with the seated opposition. I would respect him if he didn’t, but perhaps he’s able to influence party policy over time; instead of having to fight the uphill battle to get into parliament with a party of his own.

I really had no desire to get invested into this, but since this gets in my feed so frequently, here we go. I’ve read relevant posts on Daniel Berntsson’s blog, i.e. the individual which this should really concern; instead of Mullvad as a whole. I understand the donated money possibly came from Mullvad subscribers, and if that leads you to unsubscribe, by all means do so; but stop bugging the rest of us.

It seems he donated to ÖP because they criticize excessive taxation, that is believed to typically end up enriching an (upper-)middle class (the “transferiat”), that gain additional support for (self-serving) policies, from lower (working) classes that are dependent on (welfare) subsidies. And given that immigrants are typically of the latter class, it’s believed to further reinforce the transferiat’s position: by a growing dependent class voting in favor of pro-transferiat policy.

This isn’t to say he agrees with (stricter) immigration policy, nor with the other positions taken by ÖP, or any other political party for that matter. Instead, he adheres to a “libertarian anarchist” ideal, where “people can … organically and voluntarily federate together at any scale”. Rather than states appropriating capital from citizens, to only partially redistribute effectively in the form of common goods, which involves protection thereof through immigration policies for example; often to the detriment of people, including him personally.

Perhaps Cryptomator could be an option, where trust in the cloud provider isn’t required

I’ve found a video that clearly demonstrates the concept I have in mind, just to make sure we’re on the same page; because I’m not sure we are. If simple presence detections, even in the context of many vehicles in close proximity of another, and changing order through vehicles overtaking others, then I would agree; but that’s something I find hard to believe. I can’t really argue with the last point though, because it should be considered universal truth…

I understand your skepticism, but I find it difficult to believe a (team of) professional engineer(s), being unable to design a system with sufficient selectivity and resolution. And perhaps a blip is all it takes to recognize a particular vehicle, that is statistically expected to arrive at one of the further detectors within a given time frame; it really narrows down the number of possible matches when situated between two known points.

If too noisy and low resolution, it can probably not be used reliably enough, to identify vehicles purely on the basis of detected characteristics. But in the context of ALPRs at strategic locations nearby a detector, a trail of detections could be linked to a vehicle’s license plate; which would be problematic enough. If of high enough quality, the increasing reliability of ALPRs detecting vehicle make and modell, could be utilized to store characteristics under an identifier for that detected; which could allow for pattern recognition and perhaps identification in effect.

The analysis wouldn’t necessarily have to happen on-edge, but could also be offloaded to a data center; which would drive down the computation power required per unit. And yes, installing ALPRs everywhere would be more efficient, but would simultaneously be much more controversial. These coils are already in place, and the installation of the coils in the road surface doesn’t appear to be that time consuming.

Each detector seems to have two coils (see picture), which appear to be an emission coil and a reception coil; similar to metal detectors with a DD coil. Typically these coils slightly overlap to increase detection sensitivity; but this doesn’t really seem necessary when a massive block of metal hovers directly over the coils. Each coil likely consists of multiple turns of wire, that sit on top of another inside a groove cut into road surface; which should increase the inductive properties.

The magnetic field also induces Eddy currents in non-ferrous metals (meaning currents running through metal within the field), which by themselves generate opposing magnetic fields, depending primarily on the material’s conductive properties (highly conductive materials yielding stronger magnetic fields). These fields change the phase of the received signal, which in metal detectors can be used as an indicator for material properties: primarily the type of metal and its mass.

If the same were possible on the traffic detectors, or even a signature of metal concentrations, I believe you’d have a pretty robust identifier for vehicles.