Goldman Sachs estimates autonomous driving could generate an ecosystem revenue pool of about $2 trillion by 2035

Overview

In this report, Goldman Sachs continues its series on AI's impact on profit pools, using transportation as the second case study. The report notes that leading public-cloud hyperscalers' data-center capital expenditure has approached about $700 bn USD, roughly 10x the 2020 level, so the market is increasingly focused on whether AI capital investment can be converted into meaningful profit pools. The report argues that autonomous-driving commercialization has moved from proof of concept into a broader deployment phase, driven by expansion in the US, China, and EMEA, improvements in model capabilities, advances in simulation and data tools, rising consumer acceptance, and the application of end-to-end AI technologies in autonomous-driving strategies.

Core views

The core view is that autonomous driving will both create incremental demand and disrupt existing transportation markets. Goldman Sachs expects the US Robotaxi market to reach $19 bn in 2030, up from its prior estimate of $7 bn, and $48 bn in 2035; the global Robotaxi market could reach about $415 bn in 2035. Class 8 autonomous trucks are also expected to grow from a low base, with the US market reaching $16 bn in 2030 and $105 bn in 2035, while the global market could reach about $560 bn in 2035. Combining Robotaxi, consumer L3-L5 vehicles, autonomous trucks, delivery robots, software subscriptions, and digital services, autonomous-driving ecosystem-related revenue could approach $2 trillion by 2035; of this, virtual-driver technology and software-subscription revenue more directly attributable to AI would be about $300 bn.

Analysis framework

The report uses a profit-pool framework, breaking autonomous driving into Robotaxi, autonomous trucking, consumer L3-L5 vehicles, delivery robots, hardware sales, software, and digital services, and combines forecasts from regional teams and covered companies to size the global market. The analysis also compares the unit economics of autonomous versus human-driven transport, with particular focus on how declining truck hardware costs, rising driver wages, and higher vehicle utilization affect the commercialization inflection point. For potential disruption, the report estimates the scale of economic activity in US human-driven ride-hailing, taxis, truck-driver wages, and light-vehicle sales that could be replaced by autonomous driving.

Methodology notes

• market_sizingprofit_pool_analysis

  Profit pool sizing

  By estimating market revenue, gross margin, and cumulative gross profit, the report evaluates the economic value that AI capabilities may create or reallocate in the transportation industry.

• scenario_analysisbase_bear_case

  Scenario analysis

  The report uses base-case and bear-case estimates for variables such as ride-hailing share erosion, the decline in US SAAR, and VMT improvement to assess the potential impact of autonomous driving on existing markets.

• technology_assessmentautonomous_vehicle_stack

  Autonomous driving technology stack

  The report divides the autonomous-driving ecosystem into three layers—perception, decision-making, and deployment—covering sensors, chips, software, virtual drivers, Robotaxi platforms, fleet operations, and personal autonomous-driving functions.

Asset mapping & comparison

Structured mapping from thesis to named assets (strengths, weaknesses, peers, risks).

• GOOGL / Waymo

  Beneficiary of Robotaxi technology and operations

  Strengths

  Waymo has advanced fully driverless services across multiple cities, and its 30% share in San Francisco provides evidence of commercialization leadership.

  Weaknesses

  Expansion may be constrained by geofencing, regulatory approvals, vehicle supply, and operating costs.

  Comparison

  Compared with most competitors still in testing or early deployment, Waymo is closer to scaled commercial operations.

  Risks

  Tighter regulation, accident incidents, slower-than-expected cost declines, or intensified competition.

• TSLA

  Beneficiary of consumer autonomous driving and potential Robotaxi

  Strengths

  It has large-scale fleet data, an FSD roadmap, and plans for substantial AI infrastructure investment.

  Weaknesses

  The report mentions that AI infrastructure capital expenditure could reach at least $5 bn and even nearly $10 bn cumulatively, implying high capital intensity.

  Comparison

  Tesla emphasizes a generalized, less map-dependent approach, while players such as Waymo already have more evidence in geofenced commercial deployment.

  Risks

  Execution risk, regulatory risk, technology maturity, and valuation risk.

• UBER / LYFT

  Ride-hailing platforms may both benefit and be disrupted

  Strengths

  They can introduce AV supply through partnerships with autonomous-driving providers, expanding platform transaction volume and improving unit economics.

  Weaknesses

  Their existing human-driver order volume may be eroded by Robotaxi.

  Comparison

  The report's base case assumes AVs erode about 5% of UCAN ride-hailing gross bookings by 2030, versus 16% in the bear case.

  Risks

  Declining platform bargaining power, Robotaxi suppliers bypassing platforms, and slower-than-expected city expansion.

• AUR / Volvo Group

  Beneficiary of autonomous trucking and the commercial-vehicle ecosystem

  Strengths

  Autonomous trucking can benefit from rising driver wages, falling hardware costs, and higher vehicle utilization.

  Weaknesses

  The current market remains at a low base, and the commercialization path depends on safety validation, route selection, and customer adoption.

  Comparison

  Compared with Robotaxi, autonomous trucking is more B2B-oriented, and cost-per-mile advantages could become the key inflection point.

  Risks

  Regulation, insurance, accident liability, fleet procurement cycles, and slower-than-expected declines in hardware costs.

• Hesai / TEL / APTV / MBLY / Nvidia

  Suppliers of sensors, connectors, chips, software, and toolchains

  Strengths

  Expanding demand for autonomous-driving perception, computing, and system integration benefits key component and platform suppliers.

  Weaknesses

  Suppliers face price competition, technology-route changes, and substitution from customers' in-house development.

  Comparison

  The report divides the ecosystem into perception, decision-making, and deployment; suppliers sit upstream and midstream, benefiting differently from operators.

  Risks

  Sensor cost reductions compressing profits, disputes over technology routes such as lidar or maps, and vertical integration by automakers.

Key data

• Hyperscale cloud vendors' data-center capital expenditureabout $700 bn USDRoughly 10x the 2020 level, forming the backdrop for the report's discussion of AI investment returns.
• US Robotaxi market size$19 bn in 2030; $48 bn in 2035The 2030 forecast was raised from the previous $7 bn to $19 bn.
• Global Robotaxi market sizeabout $415 bn in 2035Assumes a 30%-50% gross margin for vertically integrated operators, corresponding to about $150 bn of gross profit in 2035.
• Cumulative Robotaxi gross profitabout $440 bn over the next decadeBased on global Robotaxi market expansion and gross-margin assumptions.
• US Class 8 autonomous trucking market$16 bn in 2030; $105 bn in 2035The report argues that the cost per mile of autonomous trucks in the US could become better than that of human-driven trucks in 2028.
• Global autonomous trucking marketabout $560 bn in 2035Corresponding to about $135 bn of gross profit in 2035 and about $300 bn of cumulative gross profit over the next decade.
• Total autonomous-driving ecosystem revenueabout $2 trillion in 2035Includes hardware sales, Robotaxi, autonomous trucking services, consumer autonomous-driving subscriptions, and delivery robots.
• AI-attributable autonomous-driving marketabout $300 bn in 2035Mainly includes virtual-driver technology revenue from Robotaxi and autonomous trucking, as well as consumer L3-L5 software subscriptions.
• Potentially disrupted economic scale in the USabout $440 bnCovers human-driven ride-hailing, taxis, freight, delivery-driver wages, and scenarios of declining auto sales.
• Waymo share in San Francisco30%Reached this level about 20 months after full rollout, serving as evidence of commercialization progress.

Impact & implications

The implication for capital markets is that returns on AI infrastructure investment may not only show up in cloud services or advertising, but may also form large-scale profit pools in transportation through real-world automation. Beneficiary assets may include companies with capabilities in autonomous-driving technology, chips, sensors, platform operations, fleet deployment, and software subscriptions; at the same time, traditional human-driven ride-hailing, taxi services, freight labor, and some auto sales face substitution risk. The impact on traditional automakers is more mixed: if new technology companies capture the profit share in personal mobility, they face pressure; if they can monetize through software and digital services, there is upside opportunity.

Risks

• Regulation could become stricter than expected, potentially slowing city deployment and cross-region expansion.
• Autonomous-driving safety incidents or insufficient consumer trust could weaken adoption speed.
• Industry fundamentals weaker than expected or slower-than-expected improvement in unit economics could reduce market size and margins.
• Intensified competition, elevated valuations, execution risk, and follow-on capital needs could limit the performance of some stocks.
• Traditional auto sales could be affected by shared AV substitution in the bear case, with long-term US SAAR potentially declining by 3-6 mn units.

What to watch

• The pace of city expansion by Waymo, Uber, Pony AI, WeRide, and others through the end of 2026.
• The speed at which Robotaxi commercial revenues materialize in the US, China, Europe, and other regions.
• Whether autonomous-truck hardware costs can decline from the current roughly $125K-$150K incremental cost to about $35K-$40K by 2035.
• Whether cost per mile for AV trucks in the US becomes better than that of human-driven trucks in 2028 as the report forecasts.
• Regulatory approvals, accident rates, safety data, and changes in consumer demand.
• Technological progress in Nvidia Alpamayo, simulation tools, VLA models, end-to-end autonomous driving, and map-dependent approaches.