The burgeoning autonomous vehicle (AV) industry, poised to revolutionize transportation, is facing increasing calls for enhanced accountability and transparency regarding its safety performance. Tekedra Mawakana, co-CEO of Waymo, a pioneer in self-driving technology, recently delivered a pointed message to her industry peers, emphasizing the critical need for comprehensive safety disclosures from all companies developing robotaxi technology. Her remarks, made during an interview at TechCrunch Disrupt 2025, underscore a growing debate within the sector about how best to earn and maintain public trust as these advanced systems move from testing grounds to public roads.

The Imperative of Open Data

Mawakana’s comments centered on a fundamental question of industry responsibility: who is truly committed to making roads safer? Her response was stark. "I don’t know who’s on that list, because they’re not telling us what’s happening with their fleets," she stated, highlighting a perceived lack of openness from some quarters. She articulated a clear expectation for companies that intend to remove human drivers from behind the wheel and deploy vehicles supervised remotely: "It is incumbent upon you to be transparent about what’s happening." In her view, without such transparency, companies are not "earning the right" to claim they are improving road safety.

This stance from Waymo, a company spun out of Google’s self-driving car project, comes as it actively publishes its own safety data. Waymo recently released figures asserting that its autonomous vehicles are five times safer than human drivers and twelve times safer with respect to pedestrians, based on its internal analysis. While these figures provide a benchmark, they also implicitly challenge competitors to provide comparable, independently verifiable data, thereby setting a higher standard for the industry. The comparison between human-driven and autonomous miles is complex, involving different driving conditions, operational design domains, and data collection methodologies, making direct comparisons challenging but essential for informed public discourse.

Navigating the Autonomous Landscape

The current landscape of autonomous vehicle development is characterized by intense competition and diverse technological approaches. Beyond Waymo, a handful of companies are actively pursuing robotaxi services or advanced autonomous driving systems in the U.S. These include Amazon-owned Zoox, which has recently begun offering rides in its purpose-built AVs in Las Vegas, Nevada; May Mobility and Pony.ai, both in earlier stages of commercial deployment in specific areas; and Aurora, which focuses on autonomous trucking and has published its own safety framework. Tesla, a prominent player, offers its "Autopilot" and "Full Self-Driving" (FSD) beta programs, which, despite their names, operate as advanced driver-assistance systems requiring active human supervision.

Each company operates with varying levels of autonomy and distinct operational design domains (ODDs). Waymo and Zoox, for instance, are developing Level 4 autonomous vehicles, meaning the vehicle can handle all driving tasks under specific conditions without human intervention. Tesla’s FSD, while ambitious, is generally considered Level 2, or sometimes Level 3, where the human driver remains ultimately responsible and must be prepared to intervene. This distinction is crucial for safety discussions, as the responsibility for vehicle operation shifts dramatically between these levels.

A Brief History of Autonomous Driving

The quest for self-driving vehicles has a long and storied history, predating modern computing. Early concepts emerged in science fiction, but practical experimentation began in the mid-20th century. The modern era of autonomous vehicle research gained significant momentum with the DARPA Grand Challenges in the mid-2000s, which spurred academic and industry interest in robotic navigation. Google’s self-driving car project, initiated in 2009, was a watershed moment, demonstrating the feasibility of fully autonomous operation on public roads and eventually spinning off into Waymo.

Initially, there was widespread optimism, almost a utopian vision, that fully autonomous vehicles would be ubiquitous by the early 2020s. However, the complexity of real-world driving environments, especially "edge cases" (unusual or unexpected scenarios), proved far more challenging than anticipated. The industry has since experienced a more measured evolution, moving through phases of rapid expansion, significant investment, and occasional retrenchment as the technical and regulatory hurdles became clearer. The journey from controlled test tracks to commercial deployment has been incremental, with companies often starting with limited geographic areas and specific operating conditions before attempting broader expansion.

Building Public Trust Through Transparency

Public perception and trust are paramount for the widespread adoption of autonomous vehicles. Incidents, even minor ones, involving AVs can quickly erode confidence and fuel skepticism. High-profile crashes or operational mishaps, regardless of fault, tend to attract significant media attention and often lead to public apprehension. This makes transparency not just a matter of corporate responsibility, but a strategic imperative for the industry’s long-term viability.

When companies operate with limited or opaque safety reporting, it creates an information vacuum that can be filled by speculation or fear. Transparent data, rigorously collected and openly shared, can help the public, policymakers, and regulators understand the true safety profile of these systems. It allows for informed decision-making, fosters constructive dialogue, and enables the development of appropriate regulatory frameworks. Without it, the social license to operate, which is essential for any transformative technology, becomes increasingly difficult to secure.

The Tesla Enigma: Autopilot vs. Robotaxi Aspirations

While Mawakana refrained from naming specific companies, her remarks about lack of transparency regarding "what’s happening with their fleets" strongly resonate with ongoing discussions surrounding Tesla’s approach. Tesla publishes quarterly "vehicle safety reports," which detail crash rates for vehicles using Autopilot compared to those without. However, this data primarily focuses on Level 2 driver assistance technology used predominantly on highways, where crash rates are generally lower than on surface streets. Crucially, these reports do not provide comprehensive safety data for Tesla’s nascent robotaxi pilot program in Austin, Texas, which aims for a higher level of autonomy on more complex urban roads.

The distinction between driver assistance features like Autopilot and a fully autonomous robotaxi service is critical. Autopilot requires an attentive human driver ready to take control at any moment. A robotaxi, by definition, operates without a human safety driver behind the wheel, shifting the entire burden of safety to the autonomous system. This difference in operational responsibility necessitates a vastly different standard of safety validation and reporting. The lack of detailed, publicly available data on Tesla’s full self-driving efforts, particularly its unmonitored robotaxi operations, stands in contrast to the transparency advocated by Waymo and other developers pursuing higher levels of autonomy.

Challenges in Standardizing Safety Metrics

One of the significant challenges facing the autonomous vehicle industry is the lack of universally agreed-upon, standardized safety metrics and reporting protocols. How do you definitively prove an AV is "safe"? Is it by having fewer accidents than human drivers? By operating for a certain number of miles without critical interventions? The complexity of real-world driving, combined with the statistical rarity of serious accidents, makes it difficult to draw statistically significant conclusions quickly.

Early regulatory efforts, such as California’s disengagement reports (where companies reported instances of human safety drivers taking control), provided some insight but were later criticized for their inconsistencies and limited utility as a true safety indicator. Companies now use a variety of internal metrics, including simulated miles, structured test track scenarios, and real-world operational data. The challenge lies in harmonizing these diverse datasets into a cohesive, comparable framework that can be understood by the public and used effectively by regulators like the National Highway Traffic Safety Administration (NHTSA) and state-level Department of Motor Vehicles. Developing such standards is crucial to ensure fair comparisons, encourage best practices, and build a robust safety culture across the entire industry.

The Path Forward: A Shared Responsibility

Mawakana’s call for greater transparency is not just a competitive jab; it’s a reflection of the industry’s collective responsibility to ensure that this transformative technology is deployed safely and ethically. As autonomous vehicles transition from niche experiments to a mainstream mode of transportation, the onus is on every developer to demonstrate their commitment to safety through verifiable data and open communication.

The future of autonomous mobility may hinge on whether companies choose collaboration over competitive secrecy when it comes to safety. Establishing industry-wide standards for data collection, analysis, and reporting could accelerate public acceptance and foster innovation responsibly. Ultimately, earning the public’s trust will require more than just technological prowess; it will demand an unwavering commitment to openness, accountability, and a shared vision for a safer, more efficient transportation ecosystem. The road ahead for robotaxis is long, but clarity and transparency may well be the clearest path to their successful destination.