Are Truck Drivers Safe from AI? The Autonomous Vehicle Risk Picture in 2026
Heavy truck and bus drivers score 2.5/10 on AI exposure - a low score, because driving is not a text-manipulation task and large language models offer little threat to it. Their robotics risk score is 7.5/10. The threat to the 2.1 million US heavy truck drivers (BLS OEWS May 2025, median $54,320) and 180 million globally (ILO ILOSTAT) is not chatbots. It is autonomous vehicle technology progressing through the SAE levels on fixed highway routes.
Key findings
- Truck and bus drivers (ISCO 83): AI exposure 2.5/10 - low; robotics (autonomous vehicle) risk 7.5/10 - high
- 2.1 million US heavy truck drivers (BLS OEWS May 2025, median $54,320/yr)
- Globally: approximately 180 million workers in ISCO 83 (ILO ILOSTAT 2024)
- Timeline: long-haul highway routes (80% of trucking revenue-miles) are the earliest displacement target
- Local delivery, construction, and agricultural driving score significantly lower on robotics risk (3.0-4.5/10)
Why the threat is robotics, not AI
The distinction between AI exposure and robotics risk matters most for truck drivers. AI exposure in this dataset measures susceptibility to large language models and software automation - text processing, pattern recognition, decision support. A truck driver's core task (controlling a 40-tonne vehicle through traffic, weather, and road variation) is not a text task. Chatbots do not drive trucks. Score: 2.5/10.
Robotics risk measures susceptibility to physical automation - machines that do physical work previously done by humans. Autonomous vehicles are the largest single robotics deployment targeting a specific occupation group globally. The technology is machine vision, sensor fusion (lidar, radar, cameras), and real-time path planning. It directly replaces the physical control task that truck drivers perform. Score: 7.5/10.
The AI score for truck drivers is often misread as safety. It is not. It is a category-specific score showing that language models are the wrong tool for the job - but not that the job is safe. The robotics score tells the correct story.
The SAE level framework - what stage is autonomous trucking at?
| SAE Level | Description | Status (2026) | Route type |
|---|---|---|---|
| Level 2 | Partial automation (ADAS, lane-keeping, adaptive cruise) | Universal in new trucks | All routes |
| Level 3 | Conditional automation - driver must be ready to retake control | Commercial deployment (Waymo Via, Aurora) | Highway segments |
| Level 4 | High automation - no driver required on specific routes | Limited commercial (Texas/Arizona US freight corridors) | Fixed highway routes |
| Level 5 | Full automation anywhere | Not achieved (2026) | All routes |
Aurora Innovation began commercial Level 4 driverless trucking on the Dallas-Houston freight corridor (Texas) in 2025, the first commercial deployment of fully driverless Class 8 trucks on public roads at scale. Waymo Via and Torc Robotics are on comparable timelines for other fixed US freight corridors.
The economic logic is compelling: a long-haul truck driver earns approximately $54,320 per year (BLS May 2025) and can legally drive a maximum of 11 hours per day (FMCSA Hours of Service regulations). An autonomous truck on a fixed highway route can theoretically operate 24 hours a day with no HOS limitations. At the same fuel cost, the revenue-per-mile calculus fundamentally changes.
Which truck driver jobs are safest and why
| Role | AI score | Robotics risk | US workers (BLS) | Median wage |
|---|---|---|---|---|
| Long-haul highway truck drivers | 2.5/10 | 8.5/10 | 1,400,000 | $54,320 |
| Bus drivers (fixed routes, urban) | 2.5/10 | 7.5/10 | 165,000 | $51,680 |
| Light truck / delivery van drivers | 2.5/10 | 6.0/10 | 1,550,000 | $41,700 |
| Local distribution drivers (urban) | 2.0/10 | 5.0/10 | included above | $44,200 |
| Construction dump truck operators | 2.0/10 | 4.5/10 | 280,000 | $49,400 |
| Agricultural equipment operators | 3.5/10 | 4.0/10 | 32,000 | $43,100 |
| Hazmat speciality drivers | 2.5/10 | 3.5/10 | 42,000 | $56,800 |
The robotics risk score drops sharply as route structure and predictability decrease. Highway long-haul has the highest autonomous feasibility because: the route is fixed, lane markings are clear, other vehicles behave predictably (highway conventions), and weather is navigable by current sensor fusion systems in most US climate zones. Urban last-mile delivery involves narrow streets, pedestrians, cyclists, double-parked vehicles, and building-specific access requirements that Level 4 systems cannot consistently handle in 2026.
Country-level exposure varies enormously
The autonomous vehicle threat is concentrated in countries with established highway freight infrastructure and regulatory frameworks permitting testing. In the US, truck driving is one of the most common occupations without a college degree - the American Trucking Associations estimates 3.5 million total truck drivers when including light and medium trucks. The concentration of employment in this single occupation category means national employment effects are significant.
In India, Pakistan, Bangladesh, and sub-Saharan Africa - where ISCO 83 employment is also large in absolute terms - the regulatory, infrastructure, and capital conditions for autonomous trucking at scale are 10-15 years away from the timelines seen in the US, Europe, and China. ILO ILOSTAT 2024 shows ISCO 83 employment exceeding 8% of total employment in some lower-income economies. Displacement risk in these countries is meaningfully lower in the 2026-2030 window.
China is the other high-risk geography. TuSimple, Inceptio Technology, and SmartHaul have commercial highway autonomous freight programmes on Chinese national expressways. China's manufacturing logistics volumes and highway standardisation create conditions comparable to the US for autonomous trucking viability, and with more aggressive government deployment support.
See driver occupation risk by country
Autonomous vehicle deployment timelines and driver employment concentrations vary across 206 countries. Countries with high ISCO 83 employment share face outsized impacts when deployment reaches scale.
Open the Explore Tool →Was this analysis useful?