S13.1 · Transportation, Logistics & Mobility
Asset-heavy road freight carriers navigating a 2023-2024 freight recession, with AI compressing cost lines rather than the ownership model.
US trucking generated $906B in gross freight revenue in 2024, down from $1.004T in 2023 — a freight recession running since 2023 (American Trucking Associations, "Trucking Trends 2025"). The revenue line is not the defining fact here; the ownership structure is. This is the most fragmented segment in transportation: 91.5% of carriers operate ten or fewer trucks, 99.3% run under 100 power units, and the public carriers — Knight-Swift, J.B. Hunt, Schneider, Werner — hold only low-single-digit aggregate share. What AI does in a business built this way is compress the cost lines that decide margin in a down cycle, fuel and maintenance, while leaving the asset-ownership economics that define the business exactly where they were.
Freight demand and truck capacity do not sit in the same places. Revenue concentrates on interstate lanes — the Southeast, Texas, the Southern California-to-Midwest corridors — while capacity is dispersed nationwide, and that mismatch is what keeps spot rates volatile. ATA's Freight Transportation Forecast to 2035 projects a rebound to low-single-digit annual tonnage and revenue growth through the decade. Within that, truckload and less-than-truckload are not the same margin business: truckload runs low-to-mid-single-digit operating margins against LTL's 8-12%, because LTL's terminal networks and density requirements gate entry in a way truckload's near-total fragmentation does not.
The segment sits upstream of freight brokerage and 3PLs and downstream of vehicle OEMs, fuel and energy suppliers, and commercial insurers — and those three inputs (equipment, fuel, insurance) double as both the primary cost items and the regulatory-gating surface, via FMCSA and DOT safety compliance. Capital intensity is high. Driver labor and fuel dominate the cost structure, and the assets — trucks, terminals, maintenance infrastructure — sit on the carrier's balance sheet rather than being brokered.
A trucking operator has two adjacent directions to move, and they do different jobs. Pushing vertically into brokerage is margin defense: it captures the spread a broker would otherwise take, reallocating an intermediary's take rate rather than finding new freight dollars. Geographic fleet expansion adds lane density and terminal reach — position-building, not protection.
AI's reach into this segment is narrower than its reach elsewhere in transportation. Route optimization and predictive maintenance compress fuel and maintenance — the two largest controllable costs after driver labor — but nothing in that changes who owns the truck or who bears the freight-rate cycle. Contrast freight brokerage, where AI-driven load-matching displaces the intermediary function itself. Here, AI is a cost tool layered onto an unchanged asset-ownership model, not a substitute for the asset.
Telematics- and ELD-driven route and maintenance optimization are already standard practice. The development that would actually restructure the cost base — autonomous trucking, which removes driver labor, the largest line item — sits 5-10 years from scaled deployment. Until it arrives, the segment's economics remain a function of freight-cycle timing and capital discipline. AI adoption is not the variable.