Will Self-Driving Cars Be Legal Everywhere?

Waymo, Alphabet's autonomous driving subsidiary, became the first company to operate a fully driverless commercial robotaxi service at scale. By early 2026, Waymo completes over 150,000 paid rides per week across Phoenix, San Francisco, Los Angeles, and Austin — with zero safety drivers. Waymo's 6th generation Driver system reduces hardware cost by 80% compared to Gen 5. Expansion into Atlanta, Miami, Tokyo, and other major markets is in regulatory approval stages. Waymo's cumulative driverless mileage exceeds 20 million miles — providing a safety data foundation that human drivers cannot match per-mile.

Tesla's FSD version 12+ moved to a full end-to-end neural network approach — eliminating hand-coded rules in favor of imitation learning from 1 million+ human-labeled video clips. NHTSA data shows Tesla FSD vehicles involved in 736 reported incidents as of 2024 — though accident rate per mile remains better than average human drivers on engaged FSD miles. Musk has promised full autonomy 'by end of year' repeatedly since 2016; the Cybercab (robotaxi) launched in October 2024 without a steering wheel, betting on FSD Level 4 capability delivery. FSD has the largest autonomous vehicle dataset in the world (300 billion miles of fleet data) but has yet to achieve demonstrated Level 4 reliability.

US federal AV regulations remain fragmented — there is no unified national autonomous vehicle framework. The SELF DRIVE Act (proposed) and NHTSA's AV guidance provide frameworks but lack enforcement authority. California, Arizona, and Texas have the most permissive AV operating permits, enabling Waymo and Cruise deployments. The Cruise robotaxi incident in San Francisco (October 2023) — where a Cruise vehicle pinned a pedestrian and dragged her — caused California DMV to suspend Cruise's operating permit, illustrating how single incidents can set industry-wide regulatory sentiment. International markets (Japan, Singapore, UAE) are moving faster on regulatory approval, potentially becoming first movers.

Current US legal frameworks assume a human driver as the responsible party in vehicle accidents. AV manufacturers argue their systems — as the operator — bear liability, shifting insurance models fundamentally. Traditional auto insurers lack actuarial data for AV risk pricing. The absence of clear liability assignment creates legal uncertainty that slows deployment: municipalities fear liability exposure from permitting AV operations; commercial insurers struggle to price fleet coverage; injured parties face ambiguous legal recourse paths. California, Michigan, and Texas have passed partial liability frameworks, but federal harmonization is lacking.

AAA survey data shows AV trust among US consumers peaked at 36% comfortable with riding a fully driverless vehicle (2019) and declined to 28% by 2024 following high-profile incidents involving Tesla Autopilot and the Cruise San Francisco incident. However, users who have actually ridden in Waymo vehicles show dramatically higher satisfaction and trust rates — Waymo's own data shows 94% of users feel safer in Waymo than with a human driver after their first ride. Trust appears to be a first-ride problem: once experienced, the technology converts skeptics. Mass adoption thus follows a demonstration loop rather than a public opinion shift.

Tesla's camera-only vision approach (eliminating expensive lidar sensors) directly competes with Waymo's sensor fusion approach (lidar + radar + cameras). Tesla argues its 300 billion miles of real-world fleet data makes camera-only superior; Waymo argues lidar provides 3D spatial certainty impossible from cameras alone. The resolution of this debate has major cost implications: lidar-based systems cost $50,000-$100,000+ per vehicle; camera-only could enable autonomous capabilities at $5,000-$10,000 incremental cost. If Tesla's camera-only approach achieves Level 4 reliability, it would dramatically accelerate the path to mainstream autonomous vehicles in consumer-owned cars, not just robotaxi fleets.

“Waymo's CEO Dmitri Dolgov outlined an expansion roadmap targeting 10+ US cities for driverless commercial operations by 2027, underpinned by the Gen 6 hardware cost reduction. Waymo raised $5.6 billion in external funding in 2024 (from a16z, Silver Lake, and others), providing sufficient capital for the expansion. Waymo is the only company globally with a demonstrated, commercial-scale, fully driverless Level 4 service — giving them an 18-month+ operational lead over the closest competitor.”

Quelle: Waymo (Alphabet subsidiary)

“Musk unveiled the Tesla Cybercab at a Hollywood event in October 2024, claiming Tesla would operate a robotaxi network by 2026 and eventually have 10 million Cybercabs. This prediction follows a decade of similar promises. Tesla's FSD has advanced significantly with Version 12, but independent assessments rate it at Level 2+ (driver assistance requiring human monitoring) rather than Level 4 (fully autonomous in defined operational domains). Musk's credibility on AV timelines is significantly discounted by analysts given his history of missed predictions in this domain.”

Quelle: Elon Musk, Tesla CEO

“RAND's landmark study calculated that autonomous vehicles would need to drive 275 billion miles to demonstrate — with statistical confidence — a 20% improvement over human drivers in fatality rates. At 2023 testing rates across all AV companies combined (~100 million miles/year), this would take 2,750 years. RAND argued that real-world deployment (rather than closed testing) and simulation must both contribute to safety proof, effectively endorsing the graduated deployment model that Waymo is pursuing rather than waiting for pre-deployment statistical certainty.”

Quelle: RAND Corporation

“Morgan Stanley's lead AV analyst projected the total autonomous vehicle market at $1.6T by 2030, spanning robotaxi services, autonomous trucking (highly economical for long-haul), and consumer AV hardware/software. Jonas specifically cited Waymo's unit economics improving toward profitability at scale (break-even projected at 15,000 rides/day per city, which Phoenix and SF are approaching) as evidence that commercial viability, not just technical feasibility, is on the horizon.”

Quelle: Morgan Stanley (Adam Jonas)

“GM's Cruise robotaxi subsidiary suspended all autonomous operations globally following the October 2023 San Francisco incident and regulatory sanction. GM wrote down $10B+ in Cruise investments and reduced its AV investment commitment — a significant setback for the industry. The Cruise collapse demonstrated that a single high-profile safety incident can halt an entire AV program, regardless of overall safety statistics. This regulatory risk factor is not captured in pure technology analysis and represents one of the most significant barriers to the 2028 mainstream timeline.”

Quelle: GM / Cruise