In this short article, we will describe a potential threat of the futur: the risk for autonomous cars to be weaponized by independents or state related actors.
Autonomous cars cybersecurity: Advanced vehicle technology means basically any Internet-connected car can be hacked and remotely overtaken. Risks could come from third-party devices, keyless entry, ignition control, tire pressure monitoring, or navigation, diagnostic, and entertainment systems (this article expose a 2014 view about the weaponization risk).
We won't really propose an idea, but rather start an alert to resolve a potential problem. As, the hacking of autonomous car is already possible, we can assume that it won't be really long before we see the first weaponization attempts of autonomous vehicle.
This weaponization could be done by both independent hackers or state backed cells. Given that we can draw several patterns:
Mass weaponization: the individual or the group take massively control of autonomous car for cybercrime (e.g for ransom), or political reasons (e.g create chaos in a city).
Targeted weaponization: the individual or the group attack one particular person (e.g for ransom, revenge, or physical harm), or an organization (e.g by locking all their vehicles for competitive reasons).
We will speak particularly about the case where targeted or mass weaponization is used to do physical harm. Once again we can see different ways:
Targeting the passenger: the hacked autonomous vehicle is forced to make an accident (e.g by falling into a ravine).
Targeting another driver or a pedestrian: someone outside the car (e.g rolling over a pedestrian).
Such a nightmarish weaponization by independents or state founded cells could be used to suppress enemies, dissidents, competitors, or state related personalities.
We cannot deny that it would endanger liberty and really threaten states and population.
From a legal point of view, this type of attacks would remain very hard to expose and would require hard investigations, and cyber-investigations. Nonetheless, acknowledging the low predicted prevalence of autonomous car accidents (self-driving cars could cut road deaths by 90%), suspect cases may be easier to spot.
But in my own view, it is really important to take into account this potential crime scenario — hack of the vehicle — when harm or death by autonomous car accidents occurs. We have to absolutely avoid the existence of such invisible and indirect crime networks — independents or state related.
Open to thoughts:
In the worst case, such an attack on an important government personality would not only be a matter of national security, but could also bring chaos to a territory if it result in her or his death (or being badly hurt) — without speaking of eventual diplomatic consequences.
My humble advice for head of states personalities would be to remain very careful when using autonomous vehicles, and to always keep the software up to date. I'd like to say that the car industry is probably trying its best to avoid such scenarios too.
If you want to explore the legal aspect of this subject a little more, you can see this article: Autonomous car dilemma and state sovereignty.