Why the seat still belongs to a person
Will bus drivers be replaced by AI? Not on the evidence we can see today. A transit or intercity run is driving plus a public-facing shift. The driver holds a 40-foot vehicle in mixed traffic, pulls to the curb within inches of a shelter, then turns around and explains a transfer to a rider who missed one. The two halves of the job sit in the same ten minutes.
Much of the work is physical and social at once. Deploying the ramp and securing a wheelchair takes judgment about the device, the rider and the slope of the curb. Helping an older passenger to a seat, settling a fare dispute, or calming a disruptive rider is handled in person, with no script. Those tasks do not move to software by writing better software.
The vehicle itself also needs eyes and hands. Pre-trip and post-trip inspections mean checking brakes, lights, tires and fluid levels, then reporting a fault so a mechanic can act on it. Sensors flag some of this automatically, but somebody still walks the bus. The robotics panel above rates the hardware that the physical side would need at a dexterous humanoid tier, meaning machines that move and handle objects roughly like a person. That hardware is not running routes.
What AI does, what it helps with, and what drivers keep
The tasks AI can already take on are the record-keeping and planning layer around the run: logging trips and mileage, reconciling fares, building run sheets and timetables, and flagging delays to a control center. Across this job, the share of task time AI can take without a person in the loop prints as 8%. That is paperwork and scheduling, not driving.
A larger slice of the technology’s value is assistance. Collision warnings, lane monitoring, automatic braking, camera systems and predicted arrival times all sit beside the driver rather than instead of the driver. The assisted share of task time reads 34%. Our overall coverage score for this job, the share of task time AI can handle today, is 21 out of 100; the coverage method page explains how that number is built.
What is left to people is the part riders actually see: steering in live traffic, boarding and securing passengers with mobility needs, and dealing with incidents on board. The task time that still needs a human prints as 58%. It is the core of the shift, not a leftover.
What has actually been tested
Not much, in this job. The evidence grade for parity, our read on whether AI beats a typical qualified driver, prints as D, and a D grade means there is no direct test of AI against people doing this work. We give no parity number where none has been measured. Our full scoring method sets out how the grades work, and the quality parity page covers what counts as a test.
What would settle it is concrete: published, audited results from fixed-route revenue service without a staff member aboard, covering a full year of weather and night runs, with incident rates, on-time performance and complaint data for driverless vehicles set against staffed ones on the same routes. Boarding assistance and incident handling would need to be measured too, not left out of the comparison. Until an agency publishes that, the honest answer is that it has not been tested.
When the picture could change
Most likely after 2043 (8 in 10 of our scenarios). The replacement-year method explains what that window measures and how the scenarios are drawn.
Two things could pull it earlier. Dedicated right-of-way is the big one: a busway or a closed campus loop removes most of the hard driving, and the cost panel above shows how far apart monthly tooling and a staffed shift sit. Fleet turnover is the other, since new buses arrive with more sensors fitted as standard.
Two things hold it back. Safety regulation and state licensing sit with public agencies that answer to riders and unions, so approval moves slowly and in public. And the human side of the route does not shrink with better perception software: someone has to secure a wheelchair, handle an on-board incident and answer questions, which is why agencies that have trialed automation have kept staff aboard. Capital cost matters as well; transit buses run for well over a decade before replacement.
How to stay needed as a transit driver
Lean into the tasks that sit on the human side of the split. Passenger assistance and mobility-device securement is the first. Incident handling on board, including de-escalation and knowing when to call it in, is the second. Reporting mechanical faults clearly enough that maintenance can act is the third; drivers who write a useful defect report are the ones dispatch trusts.
Two skills travel well from here. One is working alongside the assistance systems already in the cab, so alerts help rather than distract. The other is training and mentoring new hires, since agencies lose the most time to turnover in the first year.
What to do: compare your route type and duties against neighboring jobs before you plan a move.
Close work sits next door. School bus drivers share the vehicle but add student management and crossing procedures. Shuttle drivers and chauffeurs run shorter, more mapped routes. Subway and streetcar operators work in a guided right-of-way, which changes the automation question a lot. You can also see the rest of the motor vehicle operators family, the wider transportation and warehousing sector, or where driving roles land on our list of jobs most at risk. To weigh two of them side by side, use the job comparison tool.