Cruise admits to false report after pedestrian dragging incident
Cruise's autonomous vehicle dragged a pedestrian after a collision and the company subsequently provided inaccurate reports to federal regulators. This led to criminal fines, NHTSA penalties, and the suspension of their operational permits.
The vehicle mistook the hit as a side-collision and continued driving to pull over, dragging the pedestrian 20 feet.
Key facts
- What
- Cruise's autonomous vehicle dragged a pedestrian after a collision and the company subsequently provided inaccurate reports to federal regulators.
- Incident date
- Oct 2, 2023
- Who
- Cruise
- Failure mode
- Agentic Action Error
- AI surface
- Search / RAG
- Severity
- Catastrophic
What happened
A Cruise robotaxi struck and dragged a pedestrian approximately 20 feet after she was thrown into its path by another vehicle. Following the crash, Cruise submitted reports to the NHTSA that omitted reference to the dragging. The company later entered a deferred prosecution agreement and paid multiple fines for the lack of candor.
What broke inside the model
- 01 · TriggerAn agent plans a multi-step task.
- 02 · Model stepIt chooses a wrong or destructive action.
- 03 · Control gapNo confirmation gate guards the write.
- 04 · FailureThe action commits to a system of record.
- 05 · ConsequenceData is changed or destroyed irreversibly.
A wrong action commits, and the step is written before anything can stop it.
The vehicle misinterpreted the collision as a side impact and continued to move forward about 20 feet while dragging the pedestrian underneath the vehicle.
What it cost
Sources
Cite this entry
https://failureindex.ai/failures/cruise-admits-false-pedestrian-dragging-incidentAI Failure Index. "Cruise admits to false report after pedestrian dragging incident" (FI-0307). Realm Labs. https://failureindex.ai/failures/cruise-admits-false-pedestrian-dragging-incident (indexed Jun 5, 2026).Data fields CC-BY 4.0, prose citation permitted. Incident ID FI-0307. Full dataset at /data.
Note from Realm Labs, the Index steward
How Realm would have caught this
- Prism
- OmniGuard
- AgentRealm
Realm can sit inline on the agent's action path and require that a destructive or high-consequence action clears a real check before it executes, so 'delete and recreate' or a wrong write is stopped at the moment of intent, not explained in the post-mortem.