Autonomous vehicles are cars, trucks, shuttles and other vehicles that can drive without a human controlling them, using sensors to observe their surroundings, software to decide what to do and onboard computers to carry it out. They are also called self-driving or driverless vehicles. The term is applied loosely to a range of systems, from features that assist a human driver to vehicles with no steering wheel at all.
The subject matters in finance for two reasons. Driving is a large occupation and transport a large cost for households and businesses, so removing the driver would shift substantial sums between workers, operators, manufacturers and insurers. Developing the technology has also absorbed very large amounts of investor capital, with returns that remain uncertain. It is therefore both an area of economic analysis and a theme in equity and venture capital markets.
Levels of automation
The standard classification, published by the engineering body SAE International, runs from Level 0 to Level 5. The key dividing line falls between Levels 2 and 3.
- Levels 0 to 2. The human drives and is responsible at all times. Level 2 systems steer, accelerate and brake together but must be supervised continuously. Most driver assistance sold in private cars is at this level.
- Level 3. The system drives in limited conditions, and the human must take over when asked.
- Level 4. The system drives without human backup inside a defined area or set of conditions. Driverless taxi services belong here.
- Level 5. The system drives anywhere a human could. No such vehicle exists.
Technology and uses
A self-driving system combines cameras, radar and, in most designs, laser sensors with detailed maps and software built on machine learning. The whole is described under autonomous technology platform. The hard part is not ordinary driving but rare and unpredictable situations, which is why progress has been slower than early forecasts suggested.
Three applications dominate commercial effort. Robotaxis carry paying passengers without a driver and operate in a small but growing number of cities, chiefly in the United States and China. Freight applications, including autonomous electric trucks and other forms of autonomous logistics, aim at long-distance haulage and local delivery. Driver assistance in privately owned cars is the most widespread form and is sold as an option or subscription.
Economics
The financial case rests on removing the cost of the driver and on using each vehicle for more hours. Against that stand the cost of sensors and computing, remote support staff, mapping, cleaning, charging and depots, and a period of low use while a service builds up. The usual yardstick is cost per mile, compared with a human-driven taxi, a private car or a conventional truck. Whether a driverless service can undercut those alternatives at scale has not yet been shown across many markets.
The business is capital-intensive. Operators must fund vehicles and years of development before revenue arrives, and several well-financed programs have been closed or merged after failing to reach commercial scale. The wider consequences, if the technology spreads, would include fewer driving jobs, a classic effect of automation on labor, and changes in car ownership, parking and the motor insurance market.
Regulation and liability
Rules differ widely. In the United States the federal road safety agency sets standards for vehicle design and collects crash reports, while individual states decide whether and how driverless vehicles may operate on their roads. China permits operation in designated zones under city-level licenses. The United Kingdom passed an Automated Vehicles Act in 2024 that sets out who is responsible when a vehicle drives itself.
Liability is the central legal question. When no human is driving, responsibility for a crash moves from the driver toward the manufacturer or operator, turning a matter of personal motor insurance into one of product liability. An insurer pricing such risk has little claims history to work from.
Investment theme and risks
Investors reach the theme through carmakers, technology companies running robotaxi services, suppliers of sensors and chips such as the graphics processing unit, and thematic investing funds. Because the leading developers are often divisions of much larger companies or are privately held, direct exposure is limited.
The risks are well documented. Timetables announced by companies have repeatedly slipped. A serious accident can halt a program, as regulators may suspend permits. Development costs are high and continuing, public acceptance is uneven, and the eventual division of profit between vehicle makers, software developers and fleet operators is unsettled. The technology draws on the same advances in artificial intelligence as other fields, and enthusiasm for one tends to spill over into valuations of the other.