Tesla’s robotaxi plans stall in reverse

Tesla’s ambitious robotaxi network hit a speed bump in the second quarter, with paid miles driven falling significantly compared to the first three months of the year. According to a chart released by the company on Wednesday, the fleet of Model Y SUVs carrying paying passengers covered roughly 1.1 million miles in Q1, but that number dropped to around 700,000 miles in Q2 , a decline of approximately 36%.

The downturn contradicts Tesla’s recent messaging and strategic bets. For the past year, CEO Elon Musk has positioned the robotaxi fleet as the company’s future, even declaring in 2024 that Tesla would go “balls to the wall for autonomy.” The decline in robotaxi miles comes at a particularly challenging time, as Tesla’s core business also fell short of Wall Street expectations. The company’s stock tumbled more than 13% in early trading on Thursday.

At first glance, the cumulative chart appears to show steady growth in paid robotaxi rides from August 2025 through June 2026. But breaking down the numbers reveals a different story. The fleet has expanded to six cities across Texas and Florida, operating a mix of unsupervised and supervised vehicles. It is also likely that Tesla is counting paid miles driven in the San Francisco Bay Area, even though those branded robotaxis lack the state permits required for autonomous operation and rely on safety drivers. Tesla has classified that service as part of its “Robotaxi coverage.”

The slowdown also coincides with a notable admission from Musk during a conference call about Tesla’s second-quarter results. When asked why the robotaxi service is scaling so slowly, Musk explained that the company needs to “accumulate driving data that is specific to the Cybercab” , the gold, purpose-built, two-seater sedan expected to form the backbone of Tesla’s autonomous fleet , “before we can put a lot of them on the road.”

“Unlike, say, Model 3, Model Y, and our other vehicles where we’ve got a lot of vehicles on the road, millions of vehicles on the road, we don’t have that for Cybercab,” Musk said. “So we actually have to accumulate miles with Cybercabs that are retrofitted with steering wheels and acceleration and braking pedals, that kind of thing, to calibrate to the Cybercab chassis. As we are confident about that, the number of Cybercabs in cities will increase dramatically.”

This marks a shift from Tesla’s longstanding narrative that its fleet of nearly 10 million customer cars has been silently collecting data in the background to train future robotaxis, in addition to improving its Full Self-Driving driver assistance software.

On the call, Tesla executives framed the slow progress as a matter of safety caution. “Our goals are very ambitious for Robotaxi, but we do need to be cautious about causing any accidents or causing any harm to anyone,” Musk said. He added that he worries about negative media coverage triggering a regulatory crackdown. “Although there are, I think, 30 to 40,000 automotive deaths per year in the United States alone, most of those do not generate any press, you don’t really read about almost any of those. But if we injure even one person, it will be worldwide headline news, and regulators will immediately clamp down on our activities.”

Ashok Elluswamy, Tesla’s VP of AI, claimed the robotaxis have had “zero notable incidents” while driving “more than 380,000 miles” without a safety operator onboard. He did not define “notable incidents,” though he said “any reports have been of other actors impacting us when we were stationary.” However, Tesla has reported 22 crashes to the National Highway Traffic Safety Administration since launching its robotaxi trial. While most involve other cars hitting Tesla’s robotaxis, the company has reported three crashes caused by teleoperators moving the vehicles remotely, and multiple instances of the cars hitting low-speed objects like curbs, utility poles, and a tow truck’s bed.

This represents another narrative change. For years, Tesla claimed that regulatory hurdles were the main obstacle to full-scale robotaxi deployment, though it never specified what those prohibitive regulations were. Now the company says proving safety is the only thing holding it back. Despite being in early stages, Tesla chose this moment to celebrate its decision to build an autonomy stack that relies solely on cameras, unlike industry leader Waymo, which uses radar and lidar sensors.

“Historically, the so-called experts have always claimed that you need lidars, radars, HD maps, and the entire kitchen sink to drive safely. Here, we show that such is not true. You can have safe, comfortable, and affordable autonomy with just cameras,” Elluswamy said.

Both Musk and Elluswamy promised growth is coming. They noted that unsupervised miles traveled have grown roughly 10% every week since Tesla started offering them at the end of last year. “We’ll continue to scale, I think, very, very rapidly,” Musk said.