Spooky Or Plain Creepy: Using AI Self-Driving Cars As Stalkerware To Sneakily Stalk Someone
Thankfully, there is a growing effort toward AI For Good.
This latest mantra entails ways to try and make sure that the advances in AI are being applied for the overall betterment of mankind. These are assuredly laudable endeavors and reassuringly crucial that the technology underlying AI is aimed and deployed in an appropriate and assuredly positive fashion (for my coverage on the burgeoning realm of AI Ethics, see the link here).
Unfortunately, whether we like it or not, there is the ugly side of the coin too, namely the despicable AI For Bad.
Bad people will do bad things, even with the greatest advances in AI.
As an example of how some might attempt to twist goodness into badness, let’s consider the use of self-driving cars for a disturbingly real possibility of carrying out a stalking pursuit. This dovetails into a quite spooky or creepy trend of employing high-tech stalkerware.Let’s unpack that.
First, let’s be clear that in today’s world the overall notion of stalking is usually considered a form of criminal harassment or sometimes referred to as an act of criminal menace. The word “stalking” previously had a less sinister connotation, for example having been used in the relatively benign context of how to hunt for food and meant as a set of techniques for doing stealthy tracking of wild animals.
But the proverbial head-in-the-sand approach is essentially fatally flawed in that the evildoers will one way or another discover or find out how to achieve AI For Bad and then we will all be caught flatfooted by not having properly prepared for such an eventuality. All told, within reasonable limits, it tends to make sense to bring these topics into the open and thus increase awareness overall, aiming to effectively handle adverse or outright criminal behavior.
With that in mind, we can gingerly explore how a self-driving car might be used for stalking and then examine the offsetting means to mitigate or prevent such a nightmarish activity.
Here’s then the intriguing question to contemplate: Will the advent of AI-based true self-driving cars potentially open the door toward using the AI-powered vehicles to perform stalking and if so what can be done about it?
Let’s unpack the matter and see.
Understanding The Levels Of Self-Driving Cars
As a clarification, true self-driving cars are ones that the AI drives the car entirely on its own and there isn’t any human assistance during the driving task.
These driverless vehicles are considered a Level 4 and Level 5 (see my explanation at this link here), while a car that requires a human driver to co-share the driving effort is usually considered at a Level 2 or Level 3. The cars that co-share the driving task are described as being semi-autonomous, and typically contain a variety of automated add-on’s that are referred to as ADAS (Advanced Driver-Assistance Systems).
There is not yet a true self-driving car at Level 5, which we don’t yet even know if this will be possible to achieve, and nor how long it will take to get there.
Meanwhile, the Level 4 efforts are gradually trying to get some traction by undergoing very narrow and selective public roadway trials, though there is controversy over whether this testing should be allowed per se (we are all life-or-death guinea pigs in an experiment taking place on our highways and byways, some contend, see my coverage at this link here).
Since semi-autonomous cars require a human driver, the adoption of those types of cars won’t be markedly different than driving conventional vehicles, so there’s not much new per se to cover about them on this topic (though, as you’ll see in a moment, the points next made are generally applicable).
For semi-autonomous cars, it is important that the public needs to be forewarned about a disturbing aspect that’s been arising lately, namely that despite those human drivers that keep posting videos of themselves falling asleep at the wheel of a Level 2 or Level 3 car, we all need to avoid being misled into believing that the driver can take away their attention from the driving task while driving a semi-autonomous car.
You are the responsible party for the driving actions of the vehicle, regardless of how much automation might be tossed into a Level 2 or Level 3.
Self-Driving Cars And Stalking
For Level 4 and Level 5 true self-driving vehicles, there won’t be a human driver involved in the driving task.
All occupants will be passengers.
The AI is doing the driving.
By-and-large, AI-powered self-driving cars will be used to take someone or something from point A to point B. You get into a self-driving car, indicate the desired destination, and the AI driving system does the driving for you. Many are hoping that by having the AI as the driver, we’ll be able to avert the 40,000 annual fatalities and 2.3 million estimated injuries that arise due to car crashes in the United States alone.
The AI won’t drink-and-drive and will otherwise avoid the numerous human driving foibles that regrettably lead to the prevailing number of deaths and injuries in a human driving world.
One aspect about self-driving cars that is a marvel and eye-catching is the notion that there does not need to be a human driver at the wheel of the car. In fact, the steering wheel and the driving controls are likely to no longer be included since there is no need for them to be available to humans. The AI uses under-the-hood means to drive the vehicle. You’ve undoubtedly seen conventional cars that are hooked-up for AI driving and therefore the steering wheel and pedals are still present, but the future will undoubtedly have redesigned cars that will open up the interior for passenger luxury purposes.
In any case, a self-driving car does not necessarily need to have any passengers at all. A self-driving car might be transporting something such as perhaps delivering groceries from a grocery store to someone at home. Or a self-driving car might be roaming to be available whenever a ride request occurs. In short, we will eventually get used to the idea that there will be self-driving cars all around us, driving on our streets and byways, and be seemingly utterly empty. This will be entirely acceptable and actually become commonplace.
Hard to imagine, but undeniably inevitable.
Okay, overall, this is all part of the decidedly AI For Good pledged promise or a hoped-for heartwarming result of creating self-driving cars, namely that they will drive us around, there will be much fewer car crashes, and we can use the self-driving cars to even haul our stuff from place to place. But, as mentioned earlier herein, the tools of mankind can seemingly always be used to the detriment of humanity, or so it seems, given the long history of technologies that were started with the best of intentions and then later turned into mechanisms of undesirable ends.
Let’s shift gears and ponder how self-driving cars can be used in objectionable ways, including for stalking. As another prime example of the AI For Bad and self-driving cars, see my recent analysis on the use of self-driving cars as a drug mule for the illicit drug trade, at this link here.
Envision that an evildoer decides they want to try and follow someone and opts to use a self-driving car in that malevolent plot.
The person to be followed is already known by the stalker to be residing at 1234 Main Street and so the stalker instructs a self-driving car to go nearby, perhaps parking across the street at 1235 Main Street. This would be relatively easy to do, simply request that a self-driving car is needed to pick up someone at the 1235 Main Street location and that once it arrives there it should park and patiently wait for a presumed rider to get into the vehicle.
Meanwhile, once parked at the location, the sensors of the self-driving car are kept active and the engine is kept running. The various sensors on a self-driving car tend to include video cameras, radar, LIDAR, thermal imaging, ultrasonic, and so on. These devices are the eyes and ears, as it were, providing the AI with data about whatever is outside of the self-driving car. This is used to do scene analysis and figure out where the road is, where other cars are, and so on.
We already know and generally, accept the aspect that video cameras can be used for facial recognition. Increasingly, companies and governments are using facial recognition as a means of biometric identification, such as allowing entry into a restricted building due to a facial recognition scan or perhaps gaining access to your banking funds at an ATM.
I bring up this point because one facet of self-driving cars that I’ve repeatedly exhorted in my columns is the incredible and yet also disturbing capacity of self-driving cars to be a seemingly omnipresent roving eye (that’s the phrase I’ve used to describe the matter). Wherever a self-driving car goes, it presumably can record whatever it senses, and thus have a complete video recording (and the other sensory elements too) of having gone down a city street during the day, been on a neighborhood street at night, etc.
This data can be pushed up into the cloud of the automaker or self-driving tech firm or fleet owner of the self-driving car, using OTA (Over-The-Air) electronic communications to do so. Imagine all the video of a self-driving car that has been roaming all around town for an entire day, and an entire week, and an entire month, all of which captures the daily comings and goings of us all. This data could readily be examined using computer-based algorithms to try and find patterns.
For example, suppose that self-driving cars go past your office building each day. By inspecting the data, it would be possible to detect that you seem to go into your office building in the mornings, come out around lunchtime, go back into the office just after lunch, and leave the office in the evening at about 5:30 p.m. All of this was captured via the video cameras of the self-driving cars that were moving along the street upon which your office building borders.
That’s an alarmingly intrusive and privacy-invading use of that data.
There are those that point out the same data can be used for more above-board intentions. For example, suppose a realtor wants to be able to showcase a house that you might be interested in buying. Right now, there are usually one-time pictures that the realtor might take of the property, and perhaps you might look online via a search engine and see a picture that was taken several months ago. Via the treasure trove of data being collected daily by self-driving cars, you could get an up-to-date video of how that property looks today, maybe even right now.
In addition, you could see what else happens on that same street. Maybe kids routinely play in the front yard across the street from the house you are considering buying. The house next door is meanwhile getting an add-on and the construction work has been taking weeks, all of which you could watch on the video that was recorded by those self-driving cars passing by or roaming around.
You probably get the gist of where this is leading in terms of how the roving eye relates to the seedy act of stalking.
The self-driving car parked on Main Street is asked to be on the watch for the person that the stalker wants to follow. Via the onboard AI-enabled facial recognition, the AI driving system is scanning the scene and watching for the targeted individual. When the person comes out of their house across the street, the AI driving system lets the stalker know (likely conveying as such to the smartphone of the stalker).
At this juncture, the stalker could remotely command the AI to follow the person such as if the individual goes for a jog or a long walk. Or, if the person gets into a conventional car and drives away, the self-driving car could be put into the pursuit of that vehicle. And so on.
You might assume that certainly the person being followed would be suspicious of a self-driving car that has no one in it and seems to be nearby to them. As mentioned, we are going to eventually get used to seeing self-driving cars all around us, empty of anyone, and routinely driving along or parked. We are unlikely to take notice and will not give these ubiquitous self-driving cars any undue attention.
For reasons of cautious discretion, I won’t go into any further details about how the stalking could be conducted, other than to point out some other augmented tools that could be used too.
There are efforts underway to pair drones with self-driving cars, allowing a self-driving car to survey a scene in a bird’s eye view, sending the drone aloft from the roof of the car, or letting a drone land on the car. Some fast-food eateries are even considering delivering a bag of hamburgers and fries to an in-motion self-driving car that perchance is on the freeway, which is the fastest of fast-food delivery.
Another facet of self-driving cars is their likely facility to electronically communicate with each other, using V2V (vehicle-to-vehicle) electronic communications. Normally, this might be used by a self-driving car that detects debris on the roadway and opts to forewarn other nearby self-driving cars.
In theory, this same kind of capability and via the drones too could be used for self-driving cars to coordinate with each other, including doing a tag team effort of following someone and having each self-driving car insidiously inform the others.
Whoa, some might exclaim, this is like one of those wild conspiracy theory notions, for which a lot of oddball and unlikely elements would need to line-up for this to become real. Before you become overly concerned or frantic about this self-driving car-enabled stalking, numerous system-based barriers make this a lot harder to pull-off than might seem at initial glance.
The odds are that self-driving cars will be maintained and operated in a manner that would make the stalking scenario incredibly difficult to undertake by the everyday person. This includes strident system-imposed limits on being able to get the AI driving system to scan for someone, plus strict limits or prohibition on being able to follow someone, etc. Assuming that the proper kinds of system controls are going to be put in place, which is expected and might equally be mandated by regulations, the chances of doing this type of stalking are very slim.
That being said, the possibility does exist that this same approach could be utilized by a government that was in cahoots with the company running the self-driving cars involved in such an effort, a topic covered in my discussion about how some nation-states might choose to socially leverage self-driving cars (see the discussion at this link here).
Speaking of government activities, you might vaguely be aware that in 1994 the Driver’s Privacy Protection Act (DPPA) was federally passed in the U.S. and put into law the aspect that your driver’s info as stored or used by a Department of Motor Vehicles (DMV) must abide by mandated privacy guidelines and protocols. This helped in putting the kibosh on attempts to potentially stalk or track someone online in terms of at least getting their driving record. Imagine that if somehow self-driving car stalking does emerge, despite the odds against it happening, one would undoubtedly expect federal regulations would be enacted to aid in counteracting such machinations.
Conclusion
The good news is that we are only in the infancy of self-driving cars and so far it would appear that trying to use them for this type of guise is relatively infeasible. Breathe a hefty sigh of relief. Nonetheless, we would be unwise to let our guard down. We ought to be considering how those wrongdoers bent on AI For Bad might try to distort the benefits of AI-based true self-driving cars for vile deeds.
If you perchance spot a self-driving car following you down the street, know that the odds are it is not actually following you and merely happenstance that it seems to be doing so.
Post a Comment