Digital Transformation · 10 November 2026
Waymo School Bus Fix Questioned by US Federal Authorities
US federal authorities have requested more information after reports that Waymo's autonomous vehicles continued driving past stopped school buses despite a prior software recall. Wired reports the government does not accept that the problem is resolved.
What happened
US federal authorities have asked for more information about Waymo's autonomous vehicles after reports that the cars continued to drive past stopped school buses, despite the company's earlier software recall. Wired reports that the government does not accept Waymo's claim that the problem has been resolved.
Last year Waymo recalled its software after several of its self-driving cars were seen driving around stopped buses. The company has said the issue is fixed. The government's position, as Wired frames it, is that the underlying problem was not resolved by the recall.
Why it matters
This is a story about an AI system making safety-critical decisions in public, not only about a single vehicle recall. A software update is a hypothesis until it has been tested under real conditions, and this case shows how a fix can be declared complete by the operator while regulators still see the same behaviour. For organisations deploying AI in services that touch the public, the gap between internal sign-off and external verification is now a central operational risk.
The experience dimension is equally important. A stopped school bus is one of the most widely understood road rules, and children at bus stops are among the most sensitive users any transport service can encounter. When a system fails in a scenario that people instantly recognise, trust erodes faster than it would after a less visible fault, and regulators tend to respond to that loss of trust as much as to the technical defect.
The Renascence take
Most coverage will treat this as a dispute about whether Waymo's fix worked. The more useful question is what evidence would convince a sceptical public and regulator that it did, and whether that evidence was ever designed into the rollout. Driverless services are often judged on their average performance, yet people remember the single incident that involved a child and a stopped bus. Any operator in an AI-driven service should be building proof of resolution around the moments that carry the highest emotional weight, and should expect to show that proof to outsiders rather than assuming an internal recall settles the matter.
A recall is an internal statement of intent, not evidence of behaviour change. The companies that earn trust in autonomous services will be those that treat regulatory questions as part of the product experience and publish their verification data before they are asked. If your organisation has a recall-and-fix process, test it against the scenario your customers would find least forgivable, and assume the outside world will check.
Sources
This briefing was written by our Newsdesk, synthesising reporting from the outlets below. Follow the links for the original coverage.
FAQ
Questions we get on this topic
More in Digital Transformation
Stay ahead of CX
Get the signal, not the noise.
The stories shaping customer experience — plus the Journal and Experience Loom — in your inbox.


