Imagine sliding into the driver’s seat or perhaps not needing one at all and still arriving safely at your destination. This possibility – of self-driving vehicles – could be edging closer. In terms of development, analysis by Omega Law Group, finds that fully autonomous vehicles logged over 25.3 million miles and delivered an astonishing 88 % fewer potential property‑damage claims and 92 % fewer potential bodily‑injury claims compared with traditional human‑driven vehicles.
These numbers reflect real‑world operations across Phoenix, San Francisco, Los Angeles, and Austin, and they raise key questions about how law, policy, and insurance must evolve to keep pace.
The main findings are captured in the following data table:
| Metric | Value | Context / Insight |
| Property‑damage claims reduction | 88% | Waymo vs human driver benchmark (25.3 miles) |
| Bodily injury claims reduction | 92% | Same dataset |
| Injury‑involving intersection crash reduction | ≈ 96% (87‑99% CI) | Rider‑Only Waymo is comparing 56.7 miles to the human benchmark |
| Manufacturer exemption cap (Part 555) | Up to 2,500 vehicles/year | NHTSA’s limit for non‑compliant AVs |
| Consumer trust is willing to ride AVs | ~13% of U.S. drivers (2025) | Indicates public hesitancy despite safety gains |
The 88% reduction in property‑damage claims and 92% reduction in bodily‑injury claims derive from a large‑scale dataset of 25.3 million driver‑only miles recorded by Waymo in partnership with Swiss Re. In real terms, the dataset included just nine property‑damage claims and two bodily‑injury claims over that distance, compared to an expected 78 and 26, respectively, for human drivers.
Unlike generic comparisons, the human baseline was calibrated using over 500,000 claims and more than 200 billion miles of exposure, ensuring geographic and operational alignment.
The study further indicates Waymo outperformed even the latest human‑driven vehicles equipped with advanced driver assistance systems (ADAS): 86% fewer property‑damage claims and 90% fewer bodily‑injury claims.
On crash‑type specificity, a separate peer‑reviewed study of 56.7 million Rider‑Only (RO) miles through January 2025 found statistically significant reductions across all crash types when compared to human benchmarks.
In particular, injury‑involving intersection (vehicle‑to‑vehicle) crash events dropped by 96% (87‑99% confidence interval) and airbag‑deployment intersection events declined by 91% (76‑98% CI). Furthermore, vulnerable‑road‑user outcomes also improved: pedestrian injury crashes fell by ≈92%, cyclist crashes by 82%, and motorcyclist crashes by 82% in the Waymo RO metrics.
Importantly, no crash‑type group showed statistically worse performance for the AV fleet, meaning no identified scenario in the dataset where the AV rate exceeded the human benchmark.
The policy metrics layered into the table reveal complementary dimensions: the 2,500-vehicle/year cap in the Part 555 exemption rule signals a cautious regulatory rollout, while the extended 5-day incident-reporting requirement marks a change in oversight structures. These changes align with the operational expansion of AVs while drawing attention to transparency and liability oversight during transition.
Taken together, these data paint a layered story: specifically, in geofenced urban operational domains, AVs are achieving major safety‑performance improvements over human driving. However, the data also highlight that this is still an early‑stage dataset. The scope remains limited in geography and volume compared to the trillions of human‑driven miles annually.
It should be noted that, in terms of current status for autonomous vehicles, rollout remains limited, reserves remain in place, and consumer trust is still at single‑digit percentage levels.
