A massive power failure in San Francisco turned a routine afternoon of autonomous rides into an unexpected stress test for driverless technology, leaving hundreds of Waymo robotaxis frozen in place and forcing the company to pause its service citywide. The blackout knocked out traffic lights, snarled intersections and stranded riders, prompting Waymo to halt pickups while engineers and city officials worked to untangle the gridlock. The incident has quickly become a defining moment for how self-driving fleets handle real-world chaos rather than the controlled scenarios of test tracks.
How a citywide blackout turned into a robotaxi stress test
The outage that hit San Francisco was severe enough to reshape daily life across the city, and it exposed how tightly autonomous vehicles are now woven into that fabric. Power failures are not rare in California, but this one cut electricity to around 130,000 customers, darkening neighborhoods, disabling signals and disrupting basic services in a way that immediately tested every system that depends on a functioning grid. In that environment, hundreds of Waymo vehicles that normally blend into traffic suddenly became static obstacles, their hazard lights flashing as human drivers tried to navigate around them.
At its peak, the blackout affected about 30% of the city, a scale that meant the disruption was visible in almost every district and on major corridors that Waymo regularly serves. Reporting on the Outage describes traffic lights going dark, transit lines slowing or stopping and emergency crews juggling calls about both infrastructure failures and traffic chaos. In that context, the robotaxis were not just a tech story but part of a broader urban systems failure, one that raised immediate questions about how much redundancy and resilience is built into the autonomous fleets that now share the streets with everyone else.
Waymo’s decision to pause rides in San Francisco
Waymo’s leadership responded by doing something that would have been unthinkable a few years ago, when the company was still in limited pilot mode: it pulled its entire commercial service in one of its flagship markets. The Company said it had “temporarily” halted its robotaxis in San Franci after the blackout, framing the move as a safety-first response to a situation in which traffic control infrastructure was compromised and vehicles were already stuck in intersections. That pause was not just a technical adjustment, it was a public admission that the system was not yet ready to operate reliably when the city’s basic signals and power backbone failed.
In statements linked to the suspension, Waymo emphasized that it was working to restore operations only once conditions stabilized and it could be confident that its cars would not worsen congestion or block emergency routes. Coverage of how Waymo suspends service amid widespread blackout-related disruption makes clear that the company framed the pause as temporary and tied to the specific conditions created by the power loss. Yet the optics of a leading autonomous operator having to shut down an entire city at once will linger far longer than the outage itself, especially for residents who saw the cars as part of the problem rather than a neutral victim of the grid failure.
What riders and drivers saw on the streets
For people on the ground, the most striking images were not of darkened skyscrapers but of driverless cars frozen mid-journey. Social media filled with clips of Waymo vehicles stopped at intersections, some lined up bumper to bumper as if queued at an invisible barrier. One widely shared Video showed multiple cars sitting motionless with their lights flashing, while human drivers honked and tried to weave around them in the absence of working signals. For riders inside the vehicles, the experience was even more disorienting, as trips that were supposed to be seamless suddenly stopped with little clarity on how long the delay would last.
Reports described “Hundreds of Google” affiliated vehicles, specifically Waymo’s self-driving cars, stranded across San Francisco, with the company acknowledging that the power failure had disrupted its standard operations. Coverage of the San Francisco power outage highlighted how multiple videos on social media captured the same pattern: cars that could not seem to interpret intersections without functioning traffic lights, even when human drivers were cautiously negotiating right-of-way. For many residents, that visual evidence did more to shape their perception of the technology than any corporate blog post or safety report.
Traffic lights, sensors and the limits of autonomy
The blackout did more than cut power to homes and offices, it also took down a significant share of the city’s traffic control network. According to one account, the blackout also took down many of the city’s traffic lights and affected Muni mass transit, with the San Francisco Mayor publicly addressing the disruption as crews worked to restore service. That detail matters because it underscores how much autonomous systems like Waymo’s rely on predictable infrastructure, from signal timing to network connectivity, even as they tout their ability to “see” and react using onboard sensors. When those external cues vanish, the vehicles default to conservative behavior, which in this case meant stopping and staying put.
Waymo has long argued that its cars can handle complex urban environments, but the outage showed that complexity is different from outright infrastructure failure. The company said that the blackout also took down key systems that its vehicles depend on, and that it was working to resume service only after conditions normalized. In practice, that meant the cars were not just reacting to the absence of light at intersections but to a cascade of secondary effects, from confused human drivers to stalled buses and emergency vehicles. The incident highlighted a hard truth for autonomy advocates: no matter how advanced the onboard software, a robotaxi fleet is still deeply entangled with the reliability of the city around it.
Scale of the outage and wider city disruption
The power failure was not a brief flicker but a sustained event that reshaped the city’s rhythm for hours. At its height, the Outage left tens of thousands of residents without electricity, with some reports noting that blackouts hit some 125,000 homes and businesses throughout the day. That figure, alongside the separate count of 130,000 customers, illustrates both the scale and the variability of the reporting, but in either case the numbers point to a city where roughly a third of residents and commercial users were suddenly in the dark. In that environment, every traffic light that failed and every stalled vehicle became part of a larger safety challenge.
Transit systems were hit as well, with Muni mass transit affected and service reduced or rerouted while crews assessed the damage. The San Francisco Mayor addressed the situation as a citywide emergency, not just a localized inconvenience, underscoring that the blackout was a test of resilience for everything from hospitals to street-level mobility. For Waymo, operating in a San Francisco that was already struggling with grid reliability and public skepticism about robotaxis, the outage amplified existing tensions. Residents who had previously complained about autonomous cars blocking bike lanes or making awkward unprotected left turns now had fresh evidence of what happens when those vehicles meet a city in crisis.
Waymo’s communication and coordination with officials
In the middle of the chaos, Waymo tried to project an image of control and collaboration. The company said that Throughout the outage, it closely coordinated with San Francisco city officials, a claim that suggests its operations team was in direct contact with traffic managers and emergency services as the situation unfolded. That coordination likely included decisions about where to route remaining vehicles, how to safely pull cars over and when to stop accepting new ride requests. It also reflects the reality that autonomous fleets are now large enough that their behavior can meaningfully affect congestion and emergency response times.
Waymo’s public statements emphasized that it was focused on rapidly integrating lessons from the event into its systems, a nod to the company’s broader narrative that every incident is data for improvement. The description that Throughout the blackout it was in close contact with city leaders is also a reminder that autonomous operators now function as quasi-infrastructure providers, expected to coordinate like utilities when things go wrong. For San Francisco officials, that relationship cuts both ways: they gain a partner with sophisticated mapping and traffic data, but they also inherit a new class of assets that can fail in unfamiliar ways when the grid goes down.
Why the cars stalled: safety logic versus real-world frustration
From a technical perspective, the behavior that frustrated so many drivers was exactly what the vehicles were designed to do in ambiguous or unsafe conditions. When traffic lights go dark and sensor readings conflict with expectations, the safest option in most autonomous decision trees is to stop and wait for clarity. Reports described Waymo cars stalling at intersections with their lights flashing, a pattern that aligns with conservative safety logic but looks like paralysis to anyone stuck behind them. In one account, multiple vehicles were parked at a single intersection, creating a bottleneck that human drivers had to navigate without guidance.
Coverage that Waymo halts service in San Francisco as cars stall at intersections captures that tension between algorithmic caution and human impatience. For engineers, a vehicle that refuses to proceed without clear right-of-way is a success story, especially compared with a system that might roll through a dark intersection and cause a collision. For residents, however, the sight of six or more robotaxis immobilized in front of a single crosswalk felt like a failure of common sense. The blackout turned that design trade-off into a public spectacle, one that will likely influence how regulators and companies calibrate “safe enough” behavior in future updates.
Public trust, political pressure and the road ahead
The incident lands in a city where debates over robotaxis were already intense, with some residents and labor groups arguing that autonomous fleets add congestion and threaten jobs while others see them as a path to safer, more efficient streets. The sight of Waymo vehicles frozen during a crisis will give fresh ammunition to critics who argue that the technology is being deployed faster than cities can adapt. At the same time, supporters will point out that human drivers also struggled in the blackout, and that no serious injuries linked to the robotaxis have been reported in the available accounts. The truth, as usual, sits somewhere in between: the cars neither saved the day nor caused the outage, but they did reveal how fragile the promise of seamless autonomy can be when the grid fails.
For policymakers, the blackout will likely accelerate calls for clearer rules on how autonomous fleets should behave during emergencies, including requirements for remote human oversight, priority protocols for clearing intersections and coordination with agencies like Muni and the San Francisco Mayor’s office. The fact that Blackouts of this scale can strand both riders and vehicles will also feed into national conversations about infrastructure resilience, especially as more cities invite companies like Waymo to operate on their streets. I see this episode less as a verdict on autonomy and more as a preview of the messy, incremental way that new technologies collide with old systems, forcing both to adapt under pressure.
More from MorningOverview