Waymo is going pedal to the metal to make driverless taxis as commonplace as “coexist” bumper stickers, quietly zooming past rivals in the race toward an autonomous future.

The Alphabet-owned company has 1,500 robotaxis ferrying passengers around Phoenix, San Francisco, Los Angeles, Austin, and Atlanta, with plans to roll into new locales. It quietly ramped up to provide over 250,000 weekly paid trips in the US, and beat Lyft’s market share in San Francisco, making it the clear front-runner in the quest to scale rides without a driver spewing hot takes about the local mayor.

Rivals stuck in traffic

Several companies have hit potholes while chasing the driverless taxi prize. Uber abandoned its internal efforts to develop an autonomous vehicle in 2020 and partnered with Waymo in 2024. Uber also recently inked a $300 million deal with EV startup Lucid, with the aim of putting 20,000 AVs powered by driverless tech made by a startup called Nuro on the road over six years. And General Motors abandoned its AV endeavor, Cruise, which experienced significant safety stumbles and drained cash from GM.

Amazon has yet to launch a paid service, but it is testing a couple dozen of its Zoox futuristic driverless cabs—described by some riders as a “toaster on wheels.” It plans to start giving rides to the public in Las Vegas later this year.

Tesla is tailgating

Elon Musk has been pitching investors on a vision of Tesla as a robotaxi company, offering rides in its EVs that rely on cheaper means to navigate roads: cameras, in contrast to Waymo’s pricey lidar sensors—which are among the things making it a costly endeavor for Alphabet. Tesla launched a robotaxi pilot service in Austin last month and one in San Francisco this weekend. Musk claims Tesla can rapidly expand its fleet by retrofitting its existing EVs with driverless tech, and he said this week that Tesla owners will be able to rent out their vehicles for use as robotaxis sometime next year.

But so far, Tesla’s robotaxis are invite-only and operate in a more limited area of Austin than Waymo’s, with human supervisors on board. The company will need to resolve technical glitches that have caused risky situations and get the regulatory green light in more locations to catch up with Waymo.

Meanwhile, abroad…the roads are getting dominated by Chinese startups like Pony.ai and tech behemoth Baidu, which are operating or testing driverless taxis in China, the Middle East, and Europe.

The stakes are high: The global rideshare market will nearly quadruple from $123 billion in 2024 to $480 billion in 2032, Fortune Business Insights projects.—SK

Passengers in driverless cars might be less likely to clutch the roof handle with a death grip and press their feet into the floor.

Waymo recently published a peer-reviewed study suggesting that its autonomous vehicles have a lower rate of accidents than human-driven ones, with:

• 92% fewer crashes with pedestrian injuries and 82% fewer injuries to cyclists and motorcyclists.
• 96% fewer crashes involving other vehicles at intersections, which is the leading cause of injury to drivers in the US.

Though sample sizes for some accident types are too small to draw definitive conclusions from, and the results might not hold up under adverse weather conditions, the study suggests that more autonomous cars replacing human-operated ones could save lives. The National Highway Traffic Safety Administration estimates that 39,345 people died in road accidents in the US last year, with human error a factor in many crashes.

But humans may still be a problem: Best-selling author and human behavior authority Malcolm Gladwell is worried that people will act disruptively around driverless cars—like playing soccer in the street—knowing that they are programmed to stop and wait, and that, unlike humans, the cars never get distracted, drunk, or tired.

Studies are less impactful than viral videos…when it comes to public opinion, as highly publicized glitch-related incidents can make people anxious about the prospect of getting into a car with an empty driver’s seat. A recent survey showed 43% of Americans believe AVs should be illegal.—SK

The Tesla CEO is all-in on cars that don’t need a driver. And while current market leader Waymo utilizes radar and lidar, Tesla’s self-driving cars rely on cheaper, camera-only tech. That could give Tesla a big edge in the long run—if all the bugs are worked out. When might that be? Elon Musk has provided a range of answers to that question, although his vision is closer to reality than ever. Tesla robotaxis launched in Austin last month (with a human safety monitor on board, just in case), and Business Insider reported that San Francisco is the company’s next stop. Musk also announced last month that Tesla completed its first driverless delivery to a customer, without anyone in the car.—BC

A lot of effort goes into ensuring that self-driving vehicles are roadworthy, but what about the roads themselves? Our entire infrastructure was built around the assumption that people drive cars. So, how might things change once computers are behind the wheel?

The road ahead: As is usually the case, the best way to see how traffic is changing is to look at the stoplight.

Traffic lights—or robots, as some countries call them—generally have cameras and sensors that detect the presence of cars, or they operate on timers based on typical usage patterns.

Researchers at the University of Michigan say traffic could be way smoother if stoplights had access to comprehensive driving data. While those researchers focused on traditional vehicles, self-driving cars could take things to the next level.

Signal retreat: In the future, fully autonomous vehicles could communicate with each other and interpret the results in real-time, eliminating the need for traffic lights at all, according to Short Elliott Hendrickson, a collective of engineers, architects, planners, and scientists.

In that case, roads may not need as many turning lanes, and pedestrian walkways—often placed at stoplights—could be moved to places that make more sense.

But for now…road lights are changing on a smaller scale. In Nevada, California, and Germany, Mercedes-Benz added special lights on the outside of its test vehicles to let fellow drivers know the car is in semiautonomous mode. That way, if you roll up on somebody playing Candy Crush in traffic, you can rest assured that a computer is doing the driving.—BC

Waymo currently offers autonomous rides in Phoenix, San Francisco, Los Angeles, Atlanta, and Austin. Next year, the service will roll into Miami and Washington, DC. Waymo has also acknowledged conducting tests and collecting data in a number of other cities (listed on the map), but it doesn’t always publicly release a comprehensive catalog of these places.

If you do see a lidar-hatted vehicle on the road, it doesn’t mean that location figures into the company’s long-term plan:

• Waymo goes around the country to gather data on different driving conditions.
• For example, Waymo vehicles have spent some time in Michigan’s Upper Peninsula to figure out how to handle snowy conditions.—BC

Assessing responsibility in a car crash where all parties are humans can be difficult, but determining who—or what—is at fault when a self-driving vehicle is involved presents a new complication for the legal system.

But clarity is coming—a federal trial began in Miami last Monday over whether Tesla should be blamed for the 2019 death of 22-year-old Naibel Benavides after she was struck and killed by a Model S on Autopilot with a human driver present.

In this case: George Brian McGee, 48, was behind the wheel, dropped his phone, and reached down to grab it with his foot on the gas. The vehicle blew through a stop sign, striking and killing Benavides and severely injuring her boyfriend, Dillon Angulo, while they were parked legally.

The arguments in Tesla’s first federal jury trial—the company had previously settled at least four suits over fatal crashes—are pitting human vs. machine:

• The plaintiffs contend that Tesla is at least partly responsible, since its technology failed to warn the driver of an imminent crash and didn’t recognize a stop sign, plus the automatic brakes didn’t activate.
• Tesla’s defense is that McGee had been speeding at 90 miles per hour earlier that evening and was “aggressive” and “distracted” while behind the wheel.

The legal standard could make proving Tesla’s liability difficult: As The Verge reported, it’s likely an uphill climb for the plaintiffs, since the court is using Florida’s standard, which is “whether the car manufacturer exhibited a reckless disregard for human life equivalent to manslaughter by designing and marketing the vehicle.”

As for other states…few have language on the books to determine who is liable in the case of an autonomous vehicle accident. In Alabama, the human in the car is liable for any accident. But many states still do not specify who would be at fault in an accident.

Bottom line: Should Tesla lose this case, it could hurt the company’s reputation and further damage its already flagging sales, as the company has staked its future on growing its robotaxi fleet.—DL