- Over 100 Apollo Go robotaxes simultaneously halted in Wuhan, marking the first mass outage of autonomous vehicles in China.
- The incident hits as Baidu expands operations to the UAE, UK, and Switzerland, eroding confidence at a critical time.
- Baidu's lack of transparency about causes and solutions is more concerning than the technical failure itself.
- Similar events with Waymo and Pony.ai reveal a pattern of fragility in large-scale autonomous driving.
On the night of April 1, 2026, the streets of Wuhan, China, transformed into a scene of failed science fiction. Over a hundred Apollo Go robotaxis, Baidu's autonomous driving subsidiary, suddenly ground to a halt in traffic, displaying an internal message: "Driving system failure. Staff will arrive in five minutes." But no one arrived on time. One passenger reported waiting 30 minutes just to get someone on the phone after hitting the SOS button, while her vehicle remained stuck in a center lane with traffic flowing dangerously around it.
This mass outage exposes the real risks of scaling autonomous driving without adequate backup systems, impacting public safety and trust in emerging technologies.
The Unprecedented Mass Outage
This collective shutdown marks the first time a massive fleet of autonomous vehicles has experienced simultaneous paralysis in China. Baidu, which operates over 1,000 robotaxis in Wuhan alone and has completed more than 20 million rides historically, faces a critical moment. The company just launched driverless commercial operations in Abu Dhabi and Dubai, is negotiating entry into the UK and Switzerland, and has a partnership with Uber to integrate its services. The incident hits at the worst possible time: precisely as the industry tries to convince the world it's ready for massive scaling.
Technical Explanations and Consequences
Chinese media cite anonymous sources suggesting that safety self-verification systems might have detected an anomalous condition, stopping the vehicles preventively. If this theory holds, the system would have worked as designed, but with chaotic consequences. Cars stranded on highways, passengers trapped for over 90 minutes, and collisions caused by sudden braking created a scenario where the absence of serious injuries seems more like luck than design.
Wuhan's mass outage exposes that autonomous driving still doesn't justify the public trust it demands.
Troubling Patterns Across the Industry
This isn't an isolated incident. In December 2025, a power outage in San Francisco immobilized Waymo's robotaxis, forcing emergency software updates. Months earlier, in August 2025, an Apollo Go vehicle fell into a construction ditch in Chongqing, and in May of that year, a Pony.ai car caught fire in Beijing. These events reveal a clear pattern: large-scale autonomous driving hasn't yet achieved the reliability needed to justify the public trust companies are asking for.
Baidu's Opaque Response
Perhaps more concerning than the shutdown itself is the lack of transparency. Baidu hasn't publicly explained what caused the failure, how long it took to resolve, or what measures it will take to prevent future incidents. Wuhan police confirmed the event but refused to provide details about the cause. This opacity erodes confidence just as global regulators closely watch these technologies' performance.
Implications for Autonomous Mobility's Future
The Wuhan incident serves as a stark warning: scaling autonomous driving requires not just technical advances, but robust backup systems and effective emergency protocols. With companies like GLM competing in the AI space, pressure to innovate quickly can lead to sacrificing stability. The coming months will be crucial to see if Baidu and its competitors can learn from this failure or if they'll repeat similar mistakes in their race for global dominance.