Taxi Driver
Structural exposure score derived from 1 independent source (OpenAI "GPTs are GPTs" (human_rating_beta)), each normalized against its own full reference population and combined by simple average, then converted to a percentile.
We have not ingested real-world usage data for this occupation yet. We show a band only where genuine data exists, rather than estimate one.
What a taxi driver actually does
Taxi drivers transport passengers to requested destinations on an unplanned, on-demand basis, charging a fare based on a meter or fixed fee schedule rather than a pre-scheduled route. Core tasks include picking up street hails or dispatched fares, following the most efficient route to a destination, processing payment through a meter or card reader, assisting passengers with luggage, and maintaining the vehicle's cleanliness and mechanical condition between shifts.
Drivers coordinate with a dispatcher by radio, phone, or app, and are responsible for following local taxi-licensing rules, safety regulations, and completing accident or incident reports when needed.
Most taxi drivers work for an independent cab company, a municipal taxi medallion holder, or as an owner-operator leasing or owning their own vehicle, and requirements vary considerably by city and state: some jurisdictions require a specific taxi or chauffeur license on top of a standard driver's license, while others require only the standard license plus a local permit.
Job training ranges from a few days to about a year depending on the market, and most drivers enter with just a high school diploma or equivalent.
Why it reads this way
Taxi Driver scores 60/100 on this site's structural exposure measure (Low confidence — only the OpenAI pillar covers this specific O*NET code, AIOE's older occupational taxonomy has no matching entry), a comfortable-looking reading that this site is overriding downward on the strength of real, named, commercial-scale evidence the task-exposure pillars can't see.
Both structural-score pillars rate how closely a job's tasks resemble text-based, language-model work; neither one has any way to detect a physical robotaxi vehicle driving itself, the exact kind of automation actually displacing this occupation today.
Unlike bus driving, which remains protected because no commercial-scale driverless deployment exists anywhere in that segment, Waymo alone operates paid, fully driverless robotaxi service across 10 US cities today (Phoenix, San Francisco, Los Angeles, Austin, Atlanta, Miami, Dallas, Houston, San Antonio, and Orlando), completing roughly 500,000 rides per week with a stated target of one million by the end of 2026, and is actively expanding into a dozen more cities the same year.
No individual license comparable to a nuclear reactor operator's or an air traffic controller's protects a taxi driver's core task of driving a passenger from point A to point B; municipal taxi permits regulate who may operate a cab commercially, not whether a human must be behind the wheel.
The real-world labor impact is already visible, not speculative. The New York Taxi Workers Alliance has publicly opposed robotaxi expansion over direct job-loss concerns, and in Nashville, where Lyft now offers Waymo robotaxi rides through its own app, roughly half the workers at one Lyft operations depot are reportedly former drivers who've shifted to cleaning, maintaining, and repositioning the driverless vehicles instead of driving them, real, current evidence of the role changing shape rather than a hypothetical future risk.
BLS's own Bright Outlook designation and 7%+ projected growth for this occupation reflect historical ride-hailing demand growth broadly and don't yet capture this specific, geographically concentrated robotaxi disruption in the cities where it's live today.
Skills this role draws on
Range source: BLS OEWS wage data via O*NET OnLine (2025 release), SOC 53-3054.00 'Taxi Drivers'
Average source: BLS OEWS wage data for Taxi Drivers (53-3054.00) via O*NET OnLine, 2025 wage data (Annual Median Wage)
Safer, skill-adjacent careers
These aren't generic "consider retraining" suggestions. Each one shares real skill or task overlap with taxi driver work, and each one scores meaningfully higher on our structural exposure scale, with the reasoning shown below.
Similar passenger-transport skill set and customer-facing comfort, but backed by a real, current protection taxi driving lacks entirely: no commercial-scale driverless bus deployment exists anywhere, and California's 2026 rules explicitly bar autonomous heavy-duty passenger buses while allowing autonomous freight trucks and, implicitly, the smaller robotaxi vehicles already displacing taxi work.
A related transportation-industry career built around real-time coordination and safety authority rather than point-to-point driving, an environment (mixed public grade crossings, no closed guideway) that has kept every autonomous-freight precedent confined to closed private networks.
A much stronger personal-certification wall: FAA air traffic controllers hold an individually issued FAA certificate tied directly to legal authority over aircraft separation, a form of personal licensing protection nothing in taxi driving currently carries.
Taxi Driver scores 60/100 (Low structural confidence — only the OpenAI pillar covers this code, AIOE's older taxonomy has no match), a comfortable-looking raw reading this site is overriding downward: both structural-score pillars rate text/language-model task exposure and have no way to detect a physical robotaxi vehicle driving itself, exactly the automation actually displacing this occupation. No individual license protects the core driving task the way it does for nuclear reactor operators or air traffic controllers — municipal taxi permits regulate who may operate a cab commercially, not whether a human must drive. Waymo alone runs paid, fully driverless robotaxi service across 10 US cities today with roughly 500,000 rides/week (targeting 1 million by end of 2026) and is expanding into a dozen more cities in 2026. Real labor impact is already visible: the New York Taxi Workers Alliance has publicly opposed robotaxi expansion over job-loss concerns, and in Nashville, about half the workers at one Lyft robotaxi-support depot are former drivers now cleaning and repositioning vehicles instead of driving them. BLS's 7%+ Bright Outlook growth figure reflects aggregate historical ride-hailing demand and doesn't yet isolate this geographically concentrated, named, commercial-scale disruption, as of 2026-09-21. Overall confidence rated Medium despite Low structural-data confidence, given the strength and specificity of the corroborating real-world evidence.
This is a researched judgment (checklist v1.2), not a statistical measurement — it sits alongside, not instead of, the structural exposure signal above. Next scheduled review: 2027-03-21.
Sources (5)
- https://www.onetonline.org/link/summary/53-3054.00
- https://www.onetonline.org/link/localwages/53-3054.00
- https://citipen.com/waymo-robotaxi-2026-expansion-cities/
- https://www.smartcitiesdive.com/news/robotaxi-waymo-motional-zoox-tesla-uber-lyft/759895/
- https://www.irishtimes.com/business/work/2026/09/21/robotaxis-are-coming-for-important-jobs/
Data sources & methodology
Salary data: BLS OEWS wage data via O*NET OnLine (2025 release), SOC 53-3054.00 'Taxi Drivers'. Average figure sourced separately: BLS OEWS wage data for Taxi Drivers (53-3054.00) via O*NET OnLine, 2025 wage data (Annual Median Wage).
Task descriptions: Based on O*NET occupational analysis (53-3054.00).
Real-world AI usage band: Microsoft's "Working with AI" study of Bing Copilot conversations mapped to O*NET tasks (arXiv 2507.07935), and corroborating data from the Anthropic Economic Index.
Growth projections: Much faster than average (7%+, 2024-2034, Bright Outlook designation), ~22,600 projected openings — reflects aggregate ride-hailing demand growth and does not yet isolate robotaxi disruption concentrated in the specific cities where it's already commercially live (BLS Occupational Outlook Handbook / O*NET Job Outlook), based on BLS Occupational Outlook Handbook.