ZERO
DRIVER
OBSER
VATORY

Faster than
the City

In the past decade, technology has advanced rapidly, and the number of driverless vehicles in San Francisco has skyrocketed. San Francisco has introduced Cruise, Waymo, Zoox, and Tesla's robotaxis. By living in the tech capital of the world, it's impossible not to notice the evolution of these independent corporations, their takeover, and expansion. From a first-hand perspective, I have seen Waymo evolve from clunky contraptions rigged with cameras and two engineers up front to fully autonomous vehicles.

Waymo is attempting to roll out its self-driving autonomous vehicles nationally and worldwide but faces many political roadblocks. Self-driving taxis can reduce the risk of major accidents. Around 40,000 Americans die a year in road accidents, and the vast majority are from driver errors. The introduction of autonomous vehicles means consumers have more options, and some feel safer riding with a robot than with a stranger. However, Waymo's vehicles still cost more than rideshare vehicles, and because they are still in development, they can still make mistakes. Similarly, driverless cars will also lead to some drivers losing their jobs. Despite this, Waymo is expanding in fifteen U.S. metropolitan areas and counting. Many are concerned that they will "throw millions out of work to enrich tech oligarchs."¹ By sitting in the front seat of these autonomous vehicles, I will examine how independent corporations influence public policy and impact their local industry.

Waymo has expanded its fleet from around 700 to 4,000 in the past year and has around 500,000 riders weekly as of August 2026.² That expansion means that tech issues are inevitable; Waymo has issued three federal recalls since December for software problems, including one where vehicles drove into flooded roadways. In a December 2025 power outage, Waymos stopped for at least two minutes, and officials claim that over 60 had to be manually moved. The Land Use and Transportation committee had a hearing that drew a packed City Hall Chamber, with union workers and rideshare drivers rallying against Waymo over concerns about "job losses and public safety."³ Inside the hearing, emergency leaders described the vehicles' impacts on first responders. Waymo took full responsibility and outlined upcoming changes such as updates to handle power outages. Waymo's director of product management claimed that the company can now send instructions directly to their vehicles in an outage so they can respond more decisively at dark intersections. However, there are still apparent issues around Waymo's blocking traffic. During Fourth of July 2026 celebrations, dozens of Waymos blocked roadways in San Francisco, some having to be towed after running out of power.

The concept of legitimacy refers to the perceived acceptability of an actor to a population, which provides the rationale for governance. The governed people are able to decide who holds power and the degree to which they are sovereign. In Waymo's case, the company has a lot of agency; its recalls are voluntary, and they choose to self-report.⁴ On a meso-level, Waymo raises questions of whether or not local and state governments have the power to govern and restrict autonomous vehicles by anticipating their errors, or simply wait until they glitch and step in then. Bilal Mahmood, a member of the San Francisco Board of Supervisors, claimed that he had to walk an hour home after the Fourth of July blockages. The city's mayor, Daniel Lurie, also wrote to California's transportation Secretary stating that operational standards for driverless vehicles during emergencies are "no longer enough." Matthew Raifman, who leads research on autonomous vehicle policy at UC Berkeley, states that some of the so-called "edge cases" are "edge cases for autonomous vehicles, but not necessarily edge cases for transportation."⁵ He argues that Waymo's software treats these disruptions as rare when, in reality, during an event such as the Fourth of July, it's predictable. Waymo holds a significant amount of legitimacy due to permits and its safety record, but is critiqued in a few cases in which the cars were unprepared to face certain challenges.

In global politics, realism is a theory that holds that humans are intrinsically "selfish" and "competitive by nature."⁶ Realism treats the state as the primary power holder and sees corporations as actors that operate within that framework. Therefore, the state has the ability to revoke Waymo's authority. This is what the California Department of Motor Vehicles did to Cruise in 2023 after it dragged a pedestrian who was initially hit by another car.⁷ Similarly, Waymo is testing its vehicles in Minneapolis, where Minnesota decides whether to allow commercial driverless vehicles. In 2019, Governor Walz created an advisory council on connected and driverless vehicles, and its 2025 policy recommendations would have allowed these vehicles to operate in 2026; however, a coalition of labor and transportation advocates pushed back and stopped Governor Walz from advancing them. Another bill, which Our Streets MN says was led by "Waymo and its corporate lawyers," would have given the company broad operating authority and isolation from local regulations and fees, but it did not pass. Our Streets insists that local governments must keep the power to decide whether these vehicles operate in their streets, to charge fees for the commercial use of public infrastructure, and to enforce meaningful safety standards.⁸

Realism has two branches that fit this case. Offensive realists argue that actors should strive to maximize their power whenever possible, which is exactly what Waymo does with its fast growth and heavy funding. At the same time, defensive realists argue that if an actor gets too powerful, others may balance against it through alliances. Our Streets Minnesota and other coalitions act like these types of alliances. These organizations recognize that Waymo holds a lot of control, and if it seeks too much power, there may be high costs to others.⁹ However, realism has weaknesses here. It focuses on states and largely ignores what happens on smaller scales such as city councils or companies. Realism also measures power and not legitimacy, so it cannot explain why residents and officials argue over Waymo's expansion being acceptable.

On the other hand, Waymo rejects this notion that its expansion is reckless and backs it up with statistical data. Through the end of June 2026, Waymo has tracked over 270 million fully autonomous miles, which is equivalent to over 300 human lifetimes of driving. Over these miles, Waymo had 95% fewer crashes leading to serious injuries than human drivers. They also tracked an 82% reduction in injury-causing crashes compared to human drivers. The company also reports that injury crashes involving pedestrians fell by 93%, cyclists fell by 86%, and motorcyclists fell by 82%.¹⁰ When seen through this lens, the 40,000 annual deaths from road crashes by human drivers in the United States are the real problem. By attempting to account for edge cases with the slowed deployment of Waymo, more deaths could occur from human drivers. If a city attempts to stop the company from operating, Waymo has legitimacy in arguing that waiting can cost lives.

Despite these impressive statistics, some may argue that the evidence is biased. The numbers come from Waymo, which states that autonomous and human crash data use different definitions of a crash. They also state that "not all human crashes are reported."¹¹ According to EV Wire, Waymo logs "even minor contact," which sits on Waymo's site as "downloadable files."¹² This emphasizes the company's ability to create its own definition of a crash, but simultaneously highlights its transparency and full disclosure. However, a lower injury rate across millions of miles is not an acceptable answer for an emergency crew who is unable to reach their destination when blocked by standstill Waymos on the Fourth of July. Road blockages, power outages, and recalls do not lead to injury or crash statistics but influence public opinion and trust. Jonathan Morrison, the administrator of the National Highway Traffic Safety Administration, said that the agency has found a "clear pattern" of autonomous vehicles getting in the way of police, firefighters, and paramedics. He went on to state that emergency scenes are not edge cases.¹³ This is a glaring observation that skeptics raise, calling for Waymo's democratic accountability.

In terms of economic development, the debate sides less with Waymo. The company is taking over a lot of the rideshare market. According to an article from Yahoo Finance, Waymo has taken over 15% to 19% of rider spending in San Francisco, Los Angeles, and Phoenix. Gad Allon, a Wharton professor, claims that this share is large enough to have a noticeable effect on human drivers. He told Business Insider that this will not suddenly lead to drivers losing their jobs because the supply of them is flexible, but it may lead to longer waits between rides and more unpaid repositioning. Last year, data from Gridwise, a ride-hailing data platform, illustrated that driver wages dropped in Austin, LA, Phoenix, and SF, all areas in which Waymo operates. However, researchers could not say for certain that Waymo was the cause. A Lyft spokesperson stated that they believe the "future is hybrid" and as autonomous vehicles scale, so will the market. They went on to note that Lyft rides within SF's AV operating area increased by around 20% year over year. The decline in employment of human drivers will be hard to measure because drivers are independent contractors rather than employees. Katie Wells, a senior fellow at the AI Now Institute, tracked a cohort of 40 Uber drivers over five years and found that, because of autonomous vehicles on the horizon that may take their "temporary" jobs, these drivers feel less incentivized to push for better working conditions.¹⁴

Independent corporations influence public policy less by writing the rules than by shaping the evidence the rules are judged against. Nationally, Waymo expanded from 700 to 4,000 vehicles in a year. However, San Francisco officials only addressed this large spike when problems such as the December power outage or Fourth of July blockages occurred. States hold authority but primarily employ it after problems occur. For example, when California pulled Cruise's permits, it only did so after a pedestrian accident. Realism explains the government's restrictions on Waymo's power, whereas legitimacy explains why residents and officials fight over whether the expansion is acceptable in the first place. Waymo has the evidence to defend itself, though: a 95% drop in serious injuries gives it a legitimacy that arguing and edge cases do not supersede. Despite this, their safety figures are self-reported, which may lead to the evidence becoming skewed. Similarly, wage data does not allow researchers to say that Waymo was the cause of the decline in human drivers.