The Consequence Library · How records are made and graded
A driverless Cruise robotaxi struck a pedestrian who had already been thrown into its path by a hit-and-run driver, then executed a pullover manoeuvre that dragged her roughly 20 feet. Cruise omitted the dragging from its federal crash reports. That omission produced a federal criminal charge, the loss of its operating permits, and the end of the programme.
What happened
On 2 October 2023, at Market and 5th Streets in San Francisco, a human-driven vehicle struck a pedestrian and threw her into the path of a Cruise autonomous vehicle. The Cruise AV ran over her and, not detecting that she was trapped beneath it, executed a post-collision pullover that dragged her roughly 20 feet at up to 7 mph. On 24 October 2023 the California DMV suspended Cruise's driverless deployment and testing permits, stating the company had misrepresented safety information; the DMV learned of the dragging through channels other than Cruise. Cruise gave NHTSA a verbal briefing and written one-day and ten-day reports that omitted the post-collision movement, and did not correct the written record even after supplying video that showed it. On 30 September 2024 NHTSA announced a consent order with a $1.5 million civil penalty. On 14 November 2024 the U.S. Attorney's Office for the Northern District of California announced a deferred prosecution agreement in which Cruise admitted submitting a false report to influence a federal investigation and agreed to pay a $500,000 criminal fine. The chief executive resigned, nine executives were dismissed, and on 10 December 2024 GM announced it was ending the robotaxi programme.
Where control failed
Two distinct control issues are documented. First, the vehicle executed a post-collision pullover without detecting that the pedestrian remained beneath it. Second, written federal crash reports omitted the post-collision movement even though video subsequently provided to regulators showed it.
The authority question
No human was present in the vehicle to authorise the post-collision manoeuvre in real time. The reporting history separately raises the question of what review governed the company's official account of the incident before submission to regulators.
What could be proven afterward
The vehicle's own video existed and was eventually supplied, which is what made the omission provable. That contrast converted a safety incident into a criminal matter. The record also shows regulators depended on the operator for the initial account, and the California DMV did not learn of the drag from the operator at all. Self-reported incident data, unverified against independent evidence, cannot be assumed complete.
Control state, before and after
Before the consequence
No pre-action check for a person underneath the vehicle before executing a post-collision manoeuvre. Federal crash reporting depended on the operator's own account, with no independent verification channel.
After the consequence
Permits suspended October 2023. NHTSA consent order and $1.5m civil penalty September 2024. Deferred prosecution agreement and $500,000 criminal fine November 2024. Programme ended December 2024.
Sources
- Primary: U.S. Attorney's Office, N.D. Cal., "Cruise Admits to Submitting a False Report to Influence a Federal Investigation and Agrees to Pay $500,000", 14 Nov 2024 · https://www.justice.gov/usao-ndca/pr/cruise-admits-submitting-false-report-influence-federal-investigation-and-agrees-pay
- Primary: NHTSA, Consent Order on Cruise crash reporting, 30 Sep 2024 · https://www.nhtsa.gov/press-releases/consent-order-cruise-crash-reporting
- Secondary: TechCrunch, "GM's Cruise to pay $500,000 fine to DOJ, admits submitting false report", 15 Nov 2024 · https://techcrunch.com/2024/11/15/gms-cruise-to-pay-500000-fine-to-doj-admits-submitting-false-report/
- Secondary: Forbes, "GM's Robotaxi Unit Will Pay $500,000 Fine After Submitting False Crash Report", 15 Nov 2024 · https://www.forbes.com/sites/siladityaray/2024/11/15/gms-robotaxi-unit-cruise-agrees-to-pay-500000-for-submitting-false-report-after-crash/
Record history
Published 16 August 2026. Load-bearing facts re-verified against the cited sources on 16 August 2026. Corrections and material changes are appended here with their dates. To report an error in this record, write to hello@regulayer.com with the record slug and the source you believe is authoritative.
This record describes what sources establish about a consequence and the control state around it. It separates confirmed fact from source-stated cause and from architectural analysis, and it makes no claim that any control or product would have prevented the outcome. Gap codes identify a failure class, not a remedy.