Waymo Driver - self-driving

Transparency, accountability 🙈

 While Waymo produces intelligible and useful safety information, it's autonomous driving technology suffers from several important transparency and accountability limitations, including: 

• Black box decision-making. The core Waymo Driver software - the algorithms for perception, prediction, and planning - is a proprietary trade secret and is not available for public review, or even for most regulators. This limits expert and public access to and understanding of its algorithmic and other systems. In 2022, Waymo sued the California Department of Motor Vehicles (DMV) to prevent the disclosure of safety-related information contained in its driverless-deployment application, arguing that the information, which included how the company planned to handle emergencies and descriptions of crashes, constituted trade secrets.

• Raw sensor data. The vast majority of the sensor data collected (LiDAR point clouds, raw camera feeds, etc.) is closely guarded and requires legal action (subpoena or preservation order) to access after an incident. Independent forensic experts, law enforcement, or victims' lawyers face high hurdles in obtaining the data needed to perform an unbiased accident reconstruction.

• Remote assistance. Waymo uses a human "Fleet Response" team to provide contextual information to the vehicle in complex situations. The frequency and nature of these human interventions are internally tracked but not always detailed in public reports, leading to accusations that the company underreports "interventions" by remote human operators. It obscures the true extent of the system's unassisted capability and makes it hard to differentiate between a truly autonomous decision and one guided by remote human input. 

• Comparative metrics. Critics argue that while Waymo shares data, it often uses complex metrics that make direct, apples-to-apples comparison with human crash data difficult. For example, Waymo reports every fender-tap, while humans often only report crashes that involve police or insurance. 

• Incident reporting. Regulators do not require Waymo to disclose every incident involving erratic behaviour in its fleet, with many minor incidents or near-misses going unreported. Waymo has sought to keep hidden information on how it handles emergencies, unexpected vehicle behaviour, and navigational constraints.

• Inconsistent regulation. The legal and regulatory framework for assigning fault is often fragmented in the US, varying significantly between states and cities, and is not fully addressed at the federal level. The absence of clear, national AV-specific safety standards means Waymo is largely self-regulating its "safety case," which is the company's internal justification for deeming the system safe.

Risks, harms 🛑

Waymo's main safety issue is not high-speed, serious accidents caused by Waymo Driver, but rather low-speed, property-damage incidents caused by the software misjudging its surroundings. Nonetheless, the system presents a wide range of risks, and has caused real harms, including: 

• Physical safety. Incidents involving Waymo vehicles have resulted in serious injuries and at least one fatality (as of early 2025). 

• Property damage. Waymo Driver has caused collisions resulting in property damage to public infrastructure and private property by striking roadway barriers, gates, chains, utility poles, fire hydrants, and other cars. 

• Traffic disruption. Waymo vehicles frequently cause traffic jams, bottlenecks, and delays when the system encounters an "edge case" (e.g., a complex construction zone, an unusual object, or aggressive driver behaviour) and defaults to an overly cautious stop. 

• Public service disruption. Unusual Waymo behaviour or disabled vehicles require police and fire department intervention to clear traffic and manage the scene.

• Public sector revenue loss. The rise of autonomous ride-hailing such as Waymo Driver could lead to a decline in the need for human-driven parking, resulting in lost revenue from parking meters, garages, and traffic fines for municipalities, impacting city budgets.

• Over-reliance. Drivers may become over-reliant on autonomous driving systems like Waymo Driver and pay less attention to the road, resulting in accidents. There is also the risk of forgetting basic driving skills, which could be problematic if manual intervention is required.

• Privacy loss. Waymo Driver is fundamentally a data-collecting platform, which creates significant surveillance and privacy risks. Waymo vehicles are equipped with an array of sensors (LiDAR, cameras) that record the public space, including pedestrians, cyclists, and other drivers who have not consented to being filmed, in high definition continuously, far beyond the visual scope of any human driver. Thus the fleet may create a roving surveillance network that eliminates anonymity in public spaces. In addition, Waymo's Terms of Service state that it will comply with law enforcement requests for data (such as warrants and subpoenas), thereby potentially turning  the autonomous fleet into a tool for state surveillance, where the car's private data can be used against third parties or passengers in court. Equally, Waymo could expose or otherwise compromise the personal data of its customers and employees. Equally, hackers could access Waymo systems and steal personal information. 

• Legal complexity and cost. Concerns exist about who would be liable in the event of an accident involving a Waymo autonomous vehicle, and how the company's insurance policies would cover damages. For victims of a Waymo-involved crash, pursuing compensation involves a complex product liability claim against a major corporation, which is significantly more difficult, costly, and time-consuming than a standard auto negligence claim against a human driver. 

• Unprosecutability. When Waymo Driver commits a minor traffic infraction, for example by briefly driving on the wrong side of the road, or running a school bus stop arm, the system cannot be fined or charged like a human driver. This creates a two-tiered system of justice where a human would face penalties, but the company only faces a possible investigation or software recall, thereby undermining the principle of equal accountability under the law.

• Environment. Community complaints have cited noise and visual nuisance from Waymo's 24/7 autonomous operations, especially in residential areas where vehicles are waiting for fares or circling to reposition. Furthermore, Waymo vehicles spend considerable time driving without a passenger to reposition, return to base, or pick up a rider, adding to congestion, causing unnecessary wear-and-tear, and consuming energy, partially offsetting the environmental benefits of electric vehicles.

• Employment. The widespread adoption of Waymo robotaxis could result in the loss of jobs for human drivers, particularly in industries such as trucking and taxi services - which disproportionately affect working-class and minority communities.

Lawsuits ⚖️

• Jenifer Hanki v. Waymo LLC and Alphabet Inc

• Waymo LLC v. California Department of Motor Vehicles

Research, advocacy 🧮

• Search Google Scholar

Investigations, assessments, audits 👁️

• October 2025. NHTSA. Preliminary Evaluation: Failure to Stop for School Buses (PE25-013)

• May 2024. NHTSA. Preliminary Evaluation: Unexpected ADS Behavior & Collisions (PE24-016)