How Mercedes beat Tesla to become 1st to offer level 3 autonomous personal car

ByPamela Parker KGO logo

Tuesday, August 19, 2025

How Mercedes tech beats Tesla in driverless personal cars

SAN FRANCISCO (KGO) -- Did you know Mercedes is the first automaker in the U.S. to both certify and commercialize a Level 3 autonomous driving system for everyday consumers? Here's why Tesla is not quite there yet.

We all have things that help us drive, like cruise control and lane keeping assist. Those are examples of level one automation.

A DRIVERLESS TOMORROW: How Waymo went from secret Google project to dominant robotaxi company

The levels of automation are certified internationally by the Society of Automotive Engineers (SAE), which defines six levels of driving automation, ranging from 0 (no automation) to 5 (full automation). These levels describe the extent to which a vehicle can perform the dynamic driving tasks. SAE is an international organization that develops technical standards for the automotive industry.

So when a vehicle can control both steering and acceleration, these cars are considered level two autonomy cars -- like Teslas and GM's Super Cruise cars. But a human needs to be in the seat to take control at any time.

Level three is when you really start getting into self-driving, and a car is only really considered autonomous from level 3 and up.

When it comes to public perception, when you think of self-drive driving in personal cars, you think Tesla. It's like thinking of Kleenex when you think about tissues.

Mercedes became the first automaker approved to sell a level 3 autonomous system in the U.S. via its Drive Pilot. It secured approval in Nevada in early 2023, followed by California later that year. That made the system legally available -- under certain conditions -- to consumer buyers in both states.

Tesla and Mercedes-Benz are both chasing a driverless future but in very different ways. Tesla is trying to learn its way there with AI at consumer scale and minimal hardware. Mercedes is certifying its way there with tightly-bounded operating domains, redundant sensors, and formal approvals. Both paths are reshaping what "a car for the consumer market" looks like.

Tesla's playbook

Tesla could be described as AI-first and scale-driven, betting that the vast human driving data it collects will help the software improve over time by learning from real-world situations with their customers' cars on the street to feed their neural network. This, however, positions Tesla's drivers as part of the learning curve the AI needs to perfect its data for safety.

The automaker built advanced driver assistance into its mass-market cars that it then uses real-world driving data and over-the-air updates to iterate. Hardware is intentionally sparse -- no lidar, no radar, even ultrasonic sensors were removed in 2022 as the company doubled down on a pure camera "Tesla Vision" stack.

This has caused many problems, including lawsuits and deadly crashes that have arisen from either technical issues like sudden braking and navigational errors to problems caused by drivers taking their hands and eyes off the wheel, in the belief that the car can get them where they need to go without the driver needing to pay attention.

Tesla has also misleadingly named its driver assist automation in its cars 'full self-driving" or FSD. It has run into many problems with regulators and consumer rights groups as drivers believe the vehicle is capable to be fully autonomous when the system is turned on.

Tesla's regulatory woes

In California, Tesla faces regulatory pushback and permitting gaps. The California Department of Motor Vehicles has filed a lawsuit against Tesla in July 2025, claiming that the electric car manufacturer misled consumers about the self-driving capabilities of its vehicles.

The DMV pushed to suspend Tesla's sale and manufacturing of EVs in California. It's an issue that the federal government and ABC7 have been investigating since 2018.

Yet undeterred, with no full autonomy yet, Tesla launched its robotaxi pilot in Austin, Texas on June 22, 2025, using Model Y vehicles with Full Self-Driving (FSD) software. The service is strictly invite-only, limited in geography, and requires a front-seat safety monitor, not full driverless operation.

The launch was enabled by a permit from the Texas Department of Licensing and Regulation, allowing Tesla Robotaxi LLC to function as a ride-hailing service (like Uber or Lyft) until August 6, 2026. However, this does not equate to approval for fully driverless service.

A DRIVERLESS TOMORROW: Motional and Zoox take different turns in the autonomous vehicle race

Back in California, Tesla understandably has not applied for key permits required for fully autonomous ride-hailing. The state requires approval from both the California DMV (for driverless testing and deployment) and the CPUC (for AV passenger transport).

Tesla currently holds only a DMV testing permit with safety driver and a TCP (Charter Party Carrier) permit, which permits only chauffeur-driven, non-autonomous transport.

The CPUC confirmed that Tesla cannot test or carry the public in autonomous vehicles in California -- even with a human driver present -- without proper AV permits. Despite this, Tesla's CEO Elon Musk announced the launch of a ride-hailing service in the San Francisco Bay Area. It is not a fully autonomous robotaxi service, and the service utilizes human drivers, like an Uber or Lyft, but using Tesla cars.

A DRIVERLESS TOMORROW: From farming to mining, self-driving trucks are how humans do basic tasks

Mercedes Playbook

Mercedes has always focused on a certification-first and emphasizes safety-bounded systems in its quest to advance its driverless technology in consumer cars. Mercedes ships DRIVE PILOT, an SAE Level 3 system that allows hands-off and eyes-off driving, but only under strict conditions (specific divided highways, daylight, traffic parameters). It uses a conservative sensor suite (camera, radar, lidar) plus HD maps and redundancy, and it seeks state-by-state/type approval before consumer use.

The company cleared a regulatory milestone in the U.S. when it became the first automaker with a level 3 system approved for public use in the U.S. (California and Nevada).

Who crosses the finish line?

While Tesla aims to win through data volume and rapid iteration through using its drivers as an integral part of its AI's learning curve, Mercedes, by contrast, has taken a regulation and safety-first approach and seeks to win credibility through regulatory trust and a cautious rollout.

Currently, Mercedes leads the U.S. in offering legally sanctioned, consumer-ready level 3 autonomous driving, but the cars they sell are premium and at a high price point.

A DRIVERLESS TOMORROW: Self-driving trucks transport goods on Texas freeways in expansion push, could lead to job losses

Tesla remains a popular, leading and highly-visible electric vehicle brand in the U.S. that is well-known for its driver assist capabilities and loved by its followers. But with rising competition from more automakers working to level up their autonomous capabilities in the consumer vehicle market, time will tell how this race will end.

From a concept for tomorrow to the reality of today. Take a journey through innovation, failed attempts and rise of "A Driverless Tomorrow." You can watch the full ABC7 Originals series here or wherever you stream ABC7.