- Availability: Waymo began inviting riders in San Diego on September 1, 2026
- Approval: California Public Utilities Commission, August 14, 2026, covering 18 counties including San Diego. Waymo says expansion will be gradual.
- Outside evidence: IIHS and Highway Loss Data Institute study, published July 23, 2026
- Company evidence: Waymo’s own data, published September 24, 2026
- Status: neither data set includes San Diego
Before you ride, you may wonder what the evidence says about safety. Two sources matter most: an outside study from the Insurance Institute for Highway Safety (IIHS) and Waymo’s own published numbers. For the launch details, see our separate coverage of the approval and rollout.
What the independent study found
The IIHS and its affiliate, the Highway Loss Data Institute, published their analysis on July 23, 2026. Both groups are funded by auto insurers. Lead author Eric Teoh and colleagues compared Waymo’s driverless crash reports with human driving in the same four cities: San Francisco, Phoenix, Los Angeles, and Austin. The crash reports cover 2021 to 2024, about 50 million driverless miles for Waymo. The human benchmark covers about 222 billion miles.
The researchers counted only crashes a reasonable person would report to police. On that measure, Waymo had 68% fewer crashes per mile than human drivers overall, 81% fewer injury crashes, and 85% fewer single-vehicle crashes. City by city, the picture is uneven.
| City | Police-reportable crashes per mile vs. humans | Source |
|---|---|---|
| Phoenix | 76% lower | IIHS, Jul 23, 2026 |
| Los Angeles | 71% lower | IIHS, Jul 23, 2026 |
| San Francisco | 35% lower | IIHS, Jul 23, 2026 |
| Austin | 4% higher (small sample) | IIHS, Jul 23, 2026 |
What Waymo says about itself
Waymo publishes its own comparison. On September 24, 2026, it said that through the end of June 2026 it had driven more than 270 million fully autonomous miles in Atlanta, Austin, Los Angeles, Phoenix, and San Francisco. It reports 82% fewer injury-causing crashes than human drivers, 95% fewer serious-injury-or-worse crashes, and 93% fewer pedestrian injury crashes.
Waymo’s launch announcements for Denver, San Diego, and Tampa cite 94% fewer serious-injury-or-worse crashes, which likely reflects an earlier data cutoff. All of these are company figures, and San Diego is not in either data set.
Why the numbers don’t match
The IIHS and Waymo figures differ because they use different crash definitions, time periods, cities, and benchmarks. Waymo’s 95% is for serious-injury-or-worse crashes, while the IIHS 68% covers police-reportable crashes of any severity. Putting them side by side tells you little. Each answers a slightly different question.
Why It Matters to You
If you live in or visit San Diego, you can now request a Waymo ride, and the best outside evidence suggests fewer police-reportable crashes per mile than human drivers in three of the four cities studied. That is encouraging, not a guarantee for San Diego, which neither data set covers. Waymo says expansion will be gradual, so don’t expect coverage everywhere at once.
Open questions
- The IIHS study covers 2021 to 2024 and four cities. San Diego, the Ojai vehicle, and the newest software are outside it.
- It counts police-reportable crashes only. The authors note that about half of human crashes and a third of injury crashes never reach police, so the human baseline is itself an estimate.
- Austin’s result was slightly worse than humans, though the IIHS says the sample was small.
- Only Waymo reports mileage, so the study can’t compute rates for other robotaxi companies.
- The IIHS authors say the current federal crash-reporting system is inadequate for monitoring a large expansion of driverless vehicles.
- Waymo’s figures are company-reported and benchmarked against human driving in specific cities. Waymo has also issued recalls, including one involving 3,871 vehicles in freeway construction zones, as reported by Fox 5 Las Vegas.
Quick answers
Is Waymo proven safer than human drivers?
The best independent evidence points that way for crashes that reach police, in three of four cities studied. It isn’t proof for every city, vehicle, or road type, and the data is limited.
Who paid for the IIHS study?
The IIHS and HLDI are supported by auto insurers and insurance associations. They are independent of Waymo.
Has California approved Waymo for San Diego?
Yes. The California Public Utilities Commission approved the request on August 14, 2026, covering 18 counties including San Diego. Waymo says expansion will be gradual.
Sources
- IIHS: Waymo’s driverless cars crash less often than people (Jul 23, 2026)
- Waymo: From the road, September 24, 2026 (Sep 24, 2026)
- Waymo: Welcoming our first riders in Denver, San Diego, and Tampa (Sep 1, 2026)
- Engadget: Waymo receives permission to offer rides in Sacramento and San Diego (Aug 14, 2026)
- FOX5 Las Vegas: Las Vegas traffic experts weigh in on Waymo safety concerns (Sep 16, 2026)