Before Driverless Cars Scale, Cities Need to Rewrite the Rules

Autonomous vehicles are edging out of the experimental phase and into ordinary urban service. A new study argues that the hardest part of the transition may have little to do with whether the cars can drive themselves. The real test is whether cities can absorb the behavioural, spatial, economic and political consequences that follow once autonomous mobility begins to scale.

Published in the World Electric Vehicle Journal, the study "Autonomous Vehicles and the Limits of Rapid Adoption: Unintended Consequences for Urban Mobility" combines a structured scoping review with 18 expert interviews conducted between 2021 and 2025.

The study argues that implementation speed is not determined by technology alone. It emerges from a wider urban system in which governance, demand, platform power, infrastructure and public trust interact. Those interactions can create reinforcing feedback loops that slow deployment, distort sustainability gains or push cities into mobility models they did not consciously choose.

Technology Has Crossed the Threshold; Urban Governance Hasn't

The public debate around autonomous vehicles still tends to focus on technical readiness: whether perception systems are accurate, whether crash rates compare favourably with human drivers and whether operators can run safely in mixed traffic. The study accepts that these questions remain important, but argues that bounded urban deployment has already moved far enough for another problem to dominate.

City institutions are built around relatively slow-moving systems. Transport planning, land-use decisions, public procurement and liability frameworks often take years to change. Autonomous mobility, by contrast, evolves through software updates, new business models and rapidly changing operational data. One interviewee captured the mismatch bluntly: technology moves in months, while regulation moves in years.

This gap creates what the authors describe as a "pacing problem." Cities may delay decisions because they fear acting too early, yet postponement can allow private operators to define the rules in practice. Once fleets, platforms and data infrastructures are established, governments may find themselves regulating a system whose core architecture has already been locked in.

Liability is a clear example. If an autonomous vehicle causes harm, responsibility may be contested among the manufacturer, software provider, fleet operator, infrastructure manager or passenger. Until these questions are clarified, cities and companies will remain cautious about scaling. But indefinite caution also carries a cost: it can hand first movers disproportionate influence over standards, data and market design.

Roads are no longer the only governing infrastructure. Algorithms, routing systems, fleet-management platforms and proprietary data increasingly determine who gets served, where vehicles circulate and how prices are set. Without mandatory reporting and clear data-access rules, municipalities may lose the information needed to manage congestion, equity and public space.

Convenience Could Trigger a New Congestion Cycle

Autonomous vehicles could reduce crashes, improve access for non-drivers, make off-peak travel easier and turn travel time into productive time. Yet the same features that make the technology attractive can also generate more traffic.

The study identifies demand rebound as one of the most important unintended consequences. If passengers can work, rest or consume services while travelling, longer journeys become easier to tolerate. The perceived cost of distance falls. People may commute farther, make more trips or choose autonomous vehicles instead of public transport, walking or cycling.

Fleet operations add another layer. Vehicles may travel empty while repositioning, waiting for demand, reaching charging stations or moving toward more profitable areas. Even if each vehicle operates efficiently, the total system may produce more vehicle kilometres than expected.

This creates a sustainability paradox. Autonomous vehicles could reduce emissions when combined with electrification, pooling and strong public-transport integration. But a privately owned or weakly regulated system could increase traffic, energy use and urban sprawl. The environmental outcome depends less on autonomy itself than on how the service is organised.

Accessibility gains are equally complex. Older adults, people with disabilities, children, non-drivers and residents of poorly served areas could benefit significantly. But commercial platforms are unlikely to distribute service evenly without public obligations. High-demand districts and wealthier users may receive faster, cheaper and more reliable coverage, while low-income or peripheral areas remain underserved.

The study challenges the assumption that convenience automatically produces public value. It may expand mobility while also increasing congestion, weakening public transport and widening spatial inequality. Cities need to measure not only passenger uptake but also empty trips, average occupancy, modal shifts and geographic coverage.

The Car Is Becoming a Platform and the Curb the New Battleground

According to the study, autonomous vehicles should not be understood merely as cars without drivers. They may become mobile platforms for commerce, work, tourism, healthcare, hospitality and logistics.

The paper describes a "vehicle-as-venue" model in which the ride itself becomes a commercial space. Autonomous pods could function as meeting rooms, retail showrooms, mobile clinics, dining spaces or overnight accommodation. Value would come not only from transportation but from data, user attention and services delivered during the journey.

This shift could transform market power. Companies that control the platform may gain more influence than companies that manufacture the vehicle. The ability to combine user data, route information, payments, service providers and in-vehicle experiences could create powerful network effects and encourage market concentration.

For cities, the most immediate physical pressure will fall on the curb. Pick-up and drop-off activity is already difficult to manage in many urban centres. Autonomous fleets could multiply those demands while adding charging, cleaning, sensor maintenance, inspection and repair. Mobile retail and service vehicles would also require frequent stopping in high-value locations.

The study argues that curb space may become a system bottleneck. Unlike traditional parking, this demand is dynamic and continuous. If cities do not intervene, the most profitable uses could crowd out buses, deliveries, cyclists, pedestrians and lower-priority services. It means curb governance will become vital to autonomous-mobility policy. Static parking rules will be inadequate. Cities may need dynamic pricing, time-based allocation, priority hierarchies and real-time monitoring. Public transport and accessibility services may require protected access, while operators may need to disclose occupancy, dwell time and empty movement.

The broader issue is whether public space will be managed as a civic asset or monetised as a platform marketplace. Autonomous mobility could free land currently used for parking, but it could also intensify commercial competition over the street edge. The outcome will depend on who sets the rules before large-scale deployment begins.

Faster Adoption Is Not the Same as Better Mobility

Faster rollout is usually treated as evidence of innovation and competitiveness. The authors argue instead that speed is an emergent property of the whole urban system, and that moving too quickly can generate resistance, congestion, inequity and governance failure.

The findings point toward adaptive rather than one-off regulation. Cities should not simply approve autonomous services and wait for results. They need phased deployment tied to measurable outcomes: safety, congestion, empty vehicle movement, affordability, geographic coverage, public-transport integration and data transparency.

Clear liability rules are essential, but so are rules governing data ownership and access. Municipalities need enough operational visibility to regulate curb use, monitor service quality and assess whether platforms are meeting public goals. Competition authorities may also need to address market concentration before dominant operators become impossible to dislodge.

The study is exploratory, and its limitations matter. The 18 interviews are not statistically representative, the scoping review prioritises breadth over depth, and the analysis does not quantify how large the identified effects might be. Coding was conducted by a single researcher, which may introduce interpretive bias. The findings should therefore be read as a systems framework for policy and further research, not as a forecast of inevitable outcomes.

Still, the policy relevance is substantial. For developing cities, the stakes may be especially high. Autonomous services could improve access where formal public transport is weak, but limited regulatory capacity may also allow private platforms to entrench car dependence, capture data and prioritise profitable users. Importing a technology model without adapting governance could reproduce inequalities at greater speed.