Tesla

“Teslas will be in the wild with no one in them, in June, in Austin,” the automaker’s CEO, Elon Musk, declared on Tesla’s 2024 Q4 earnings call. “This is not some far-off mythical situation; it’s five, six months away.”

While only a few words in those sentences turned out to be true, they were closer to reality than Musk’s other promises about Tesla’s self-driving ambitions in 2025. There was also the time he swore “half of the population of the U.S.” would have Robotaxi access by the year’s end. Or the many times he said that the safety driver, the human in the passenger’s seat tasked with monitoring the car as it drives along, would no longer be needed. Or the time Musk said Tesla’s Robotaxi service would operate in eight to 10 major metro areas by the end of 2025.

Any way you want to look at it, Tesla’s Robotaxi army failed to conquer the world in 2025.

The idea worked like this: Tesla would be able to deploy Model Y Robotaxis in Austin and then throughout the U.S. They’d be very similar to any Model Y you can buy right now, but with a slightly more advanced Full Self-Driving (FSD) system—the name for Tesla’s camera-based autonomous driving setup. Riders could pay to hail the taxis, and while a human safety operator would be minding things from the passenger’s seat, the car eventually wouldn’t need that at all. And over time, using AI training, the Tesla Robotaxi fleet and consumer FSD would achieve full autonomy.

If that happens, it won’t be in 2025. Despite Musk’s continual promises and predictions, the automaker’s quest to solve self-driving by the end of this year—whether it be its own private fleet of autonomously-operated robotaxis or private autonomy—has come and gone again.

To Musk’s credit, Tesla did launch its Robotaxi service in Austin in June. However, it wasn’t “with no one in them” as he claimed. Even today, safety riders remain with their fingers hovering over an emergency kill switch hidden in the door handle, despite Tesla trialing out driverless rides recently. The safety driver should be gone soon, of course—just as Musk said in September, October and earlier this month.

Granted, Musk said was able to be chauffeured around without a safety driver last week, and a few Cybercabs—the company’s two-seater taxis that supposedly don’t need a steering wheel or pedals—have been seen zooming around downtown Austin. But the Waymo-like scale that Musk promised will have to arrive in 2026, or later. 

Today, Tesla’s Robotaxi service operates in Austin and San Francisco. Musk has said that the service would expand to Dallas, Houston, Phoenix, Miami and Las Vegas. Combined, that vision represents about 15.25 million people, or around 4.5% of the U.S. population. 

From the outside, the Robotaxi launch in Austin seems massive. Thanks to the hard work of influencers everywhere, it might seem like the service is as robust as Waymo, but that’s just not the case on the ground.

InsideEVs Editor-in-Chief Patrick George visited Austin last week and got to experience the current state of the fleet first-hand. He told me that wait times for a car downtown often hovered between 15 and 25 minutes, and that’s if he wasn’t presented with a “high-demand” ride denial. Visually, Waymo’s autonomous taxis outnumber Tesla’s Robotaxis by a massive margin. And a recent analysis cited by Electrek indicated that only about three dozen Robotaxis may be operating in the city right now.

Musk said in October that Tesla would expand its Austin fleet to 500 cars by the end of 2025 and more than 1,000 cars in the Bay Area. “We’re scaling up the number of cars,” he said. “Probably we’ll have a thousand cars or more in the Bay Area by the end of this year, probably 500 or more in the greater Austin area.”

Meanwhile, the Bay Area has fewer than 150 cars. So why did Musk claim that the Robotaxi fleet would “double” in November when it had not yet even reached 250 vehicles?

FSD isn’t where Tesla claimed it would be, either. During that same second-quarter earnings call, Musk predicted that Tesla would make unsupervised FSD available by the end of the year. We’ve heard that one before, and while Musk claims you can now text and drive (which is illegal in most states), it’s still a far cry from unsupervised FSD—or, you know, that hands-free coast-to-coast drive it promised back in 2017.

The point here is, once again, that vehicle autonomy is a lot more complicated than any automaker has made it out to be.

Whether it be the feat of personal autonomy or commercial, no company has been able to deliver on full autonomy with an unfettered operational design domain. There are limits, and being open about those limits is key to earning the public’s trust.

Musk’s promises are important to investors. The company’s stock, riding high on autonomy promises, has been trending at its most valuable price point ever in recent months. Tesla needs to eventually make good on its promises before investors get tired of missed deadlines, or it needs to be more forthcoming about the difficulties and timeline to launch.

With just hours left on the clock, 2025 was not it. Maybe 2026 will be better.