I have thought about this a bunch of times and I always get hung on the potential for abuse. Here’s a few:
- Trolling; the base level of abuse, simply hitting the “jerk” button for every car, or every N% of cars at random, polluting the data.
- Character assassination; a targeted abuse of a person via reporting them as a bad driver. If your “jerk drivers” app partners with or vendors to e.g. insurance companies to determine premiums, or police to do actual enforcement, or rental companies, or… that could have very real, financial or social consequences on people.
- Neighbourhood assassination; falsely reporting jerk drivers whenever you pass through a particular neighbourhood or area. Related: reporting bad drivers abusively on university and industrial/technology campuses (think Google) to craft a reputation for the company/university/organisation as having a culture of bad or even dangerous driving.
- Fakes; abusing the technology itself to report bad driving from anywhere about any car, given you can just, for example, write out the plate number on a piece of cardboard and place it in front of the camera. Or onto a screen for more automation.
- Information leaks; if the service leaks out information like places where bad driving was reported, or even just leaking plate numbers from an API, that kind of thing. That could be used for nefarious purposes like stalking or abusive tracking.
- Slowing down traffic; to take the example of “warning other drivers of dangerous drivers” you could abuse the system by reporting lots of people driving by e.g. your house towards a motorway, causing people on the motorway to drive “more carefully” or actively avoid random cars for no actual reason, which could cause traffic disruptions.
Many countries already have a police service to report bad drivers. New Zealand has Roadwatch, UK has (I think) a service of the DVLA, etc. I would be wary of a less accountable third party that does not address abuse concerns upfront / treats them as afterthoughts.