The Human Benchmark: How a New Framework Aims to Insure Our Robot Drivers

NEW YORK, NY – June 02, 2026 – The long-promised future of autonomous vehicles (AVs) has always hinged on a single, deceptively simple question: are they safer than us? Today, insurance giant Lockton and real-world intelligence platform Nexar have unveiled a new framework that aims to provide the first truly independent answer, potentially breaking a years-long logjam that has stalled the industry's growth and kept insurers on the sidelines.

The Trust Deficit: Why a 'Human Benchmark' Matters

For years, the AV industry has been operating in a data vacuum. While developers like Waymo and Cruise publish their own safety reports detailing billions of simulated and millions of real-world miles, a lack of a universal, third-party standard has created what the new partners call "structural challenges" for the entire ecosystem. Regulators have struggled to assess competing safety claims, insurers have been forced into "blunt underwriting approaches" or have simply restricted capacity, and the public has been left to parse dueling statistics, breeding skepticism. The core problem is one of translation: how do you compare a machine's performance, often in controlled environments, to the messy, unpredictable reality of human driving?

This new framework, built on a "human-benchmark," is designed to cut through that noise. It seeks to provide a common language for safety that all stakeholders—developers, insurers, and regulators—can understand and trust. The core value proposition is not just about measuring miles driven, but about contextualizing performance against the chaotic reality of the open road. It addresses the fundamental need for an independent arbiter to validate the safety claims that are critical for mass adoption.

Under the Hood: Real-World Data vs. The Simulation

The engine behind this new standard is Nexar's massive and unique dataset. The framework is powered by BADAS 2.0, a collision anticipation model trained not on clean, simulated environments, but on 10 billion miles of unfiltered, real-world driving. This data, captured by a network of 350,000 cameras across 94% of U.S. roads, includes a classified library of 60 million safety-critical events—the near-misses, the sudden stops, and the unexpected hazards that define daily driving but are nearly impossible to simulate at scale.

This "ground truth" approach directly addresses what industry experts call the "miles-to-confidence" gap, where simulations fail to capture the long tail of real-world edge cases. The framework includes two key components: the Nexar Risk Index, which assesses the inherent risk of a specific driving environment based on geography and conditions, and Nexar Apex, a platform where developers can submit their systems for verification against curated, real-world scenarios.

"For the first time, AV systems can be benchmarked against a standard that wasn't built by anyone with a stake in the outcome," said Jon Miller, Chief Business Officer at Nexar. "That is what credible verification requires, and it is what this framework makes possible."

Unlocking the Market: A Framework for Insurers and Regulators

From a financial and industrial transformation perspective, this is where the story gets interesting. The inability to accurately price risk has been a major brake on the commercial scaling of AVs. Lockton, the world's largest privately held insurance brokerage, is not just a partner in this venture; it represents the end-user whose problems are being solved. By creating a standardized way to differentiate risk between various AV systems, the framework could finally unlock the insurance market for autonomous fleets.

"For autonomous vehicles to scale, safety has to be measured against a credible, independent benchmark built on real-world driving. This framework provides that benchmark," stated Preet Gill, EVP and Leader of Lockton's Global Technology Risk Practice. "As autonomy is adopted across logistics, mobility, and commercial fleets, even insurers who don't directly underwrite AVs will carry exposure, and our role is to help evaluate that risk before it becomes systemic."

Insurers could move from broad, conservative policies to finely tuned products based on verifiable performance data. This has massive ripple effects for logistics, mobility-as-a-service, and commercial fleet operators, for whom insurance costs are a critical operational expense. Furthermore, it provides regulators like the National Highway Traffic Safety Administration (NHTSA)—which is actively working towards a single national AV safety standard—with a powerful, data-driven tool to inform policy and avoid a harmful patchwork of state-level rules.

A Crowded Field of Standards

The Lockton-Nexar initiative does not enter an empty arena. The AV safety landscape is populated with existing standards like UL 4600, which focuses on a holistic safety case, and ISO 26262, the bedrock of automotive functional safety. Tech giants have also proposed their own solutions, most notably Mobileye's Responsibility Sensitive Safety (RSS) model, which uses a formal mathematical model to define safe driving. However, the new framework's unique selling proposition is its staunch independence and its foundation in a massive, continuously updated library of human driving behavior.

While other standards provide crucial guidelines for system design and functional safety, this framework is designed to answer the comparative question: in this specific, real-world scenario, did the AV perform better or worse than a typical human driver? This shifts the focus from theoretical safety cases, which can be difficult for non-specialists to interpret, to demonstrated, empirical performance that is instantly understandable.

The Road to Adoption

The ultimate success of the framework will depend on its adoption by AV developers themselves. The proposition is compelling: a third-party-validated safety rating could streamline regulatory approvals, lower insurance premiums, and, most importantly, build public trust. The press release notes that Nexar's data is already in use by leading AV developers to support published safety cases, suggesting a degree of industry buy-in. However, challenges remain. Developers have invested billions in their own proprietary testing and validation methodologies and may be hesitant to submit sensitive system data to a third-party platform, despite assurances of confidentiality.

The "not invented here" syndrome is a powerful force in Silicon Valley, and the cost and engineering effort required for integration could also be hurdles, especially for smaller players. Yet, the pressure from insurers and regulators for a common, verifiable standard may prove irresistible, creating a powerful market-driven incentive for developers to get on board. This initiative represents a critical test of whether a data-driven, independent standard can finally pave the road for the mass adoption of autonomous technology.