A video of a Xiaomi SU7 sedan seemingly taking itself for a spin went viral over the weekend in China, sparking safety concerns. The internet was quick to imagine hacked cars and rogue AI, but Xiaomi quickly shut down the drama, claiming that the mysterious drive was simply triggered by a remote command from the owner’s Apple iPhone.

More: Xiaomi Boss Admits They Bought Teslas And Tore Them Down To Learn Every Secret

The incident took place on September 30, when the owner, identified as Li Xiaoshuang, parked his light blue SU7 outside a store in the city of Weihai. Minutes later, while he was inside the shop with a woman, the car began to creep forward with no one behind the wheel. Realizing what was happening, Li rushed outside and sprinted after his runaway EV.

A Confusing Start

After sharing the surveillance footage, the owner contacted Xiaomi’s customer service to report the car’s strange behavior. The representative informed him that the EV had received a remote command from a device labeled “iPhone 16.2.” The owner, however, insisted he hadn’t touched his phone nor did he own an iPhone 16 and provided the video as proof.

The mix-up was later clarified when engineers confirmed that “16.2” wasn’t referring to a newer iPhone model at all, but to the internal identifier Apple assigns to the owner’s iPhone 15 Pro Max. Likewise, an iPhone 16 Pro also linked to the vehicle, belonging to the woman seen in the video, appeared under the code “17.1.”

Technical Verification

After reviewing the EV’s data logs and the smartphones’ activity records, Xiaomi engineers confirmed that the car had, in fact, received a valid remote parking command from the owner’s iPhone at that exact moment. The company also verified that every system operated correctly.

More: China Recalling Over A Third Of All Xiaomi SU7s After Deadly Crash

The company issued a formal apology for the misunderstanding and thanked the owner for cooperating in the investigation. They also emphasized that all vehicle systems operated as intended. The owner eventually acknowledged Xiaomi’s findings after witnessing the data verification himself, and urged others to stop spreading false claims about software bugs or hacks.

Accidental Commands, Real Consequences

So what could have triggered the remote parking command without the owner realizing it? The most likely explanation is that the feature was unintentionally activated through the iPhone app or its voice assistant.

Still, the idea that a simple tap or phrase can send a two-ton machine rolling on its own should give manufacturers pause. As cars become more like smartphones on wheels, it might be time for automakers to impose tighter safeguards on how these systems talk to each other, rather than assuming every “smart” feature is foolproof.

Below you’ll find the official statements from both Xiaomi and the owner.

---

Xiaomi’s Official Statement

Regarding the recent online video of a Xiaomi car suddenly driving away, we attach great importance to this incident and immediately established a special task force to investigate and collaborate with the user to verify the situation.

Both parties verified the user’s authorized mobile app operation logs and vehicle data, reaching a consensus: “The vehicle’s backend data matches the iPhone 15 Pro Max’s operation logs, response times, and vehicle exit commands, ruling out any vehicle quality issues.”

During the investigation, with the user’s consent, we obtained vehicle backend data and operation logs from two phones with vehicle control permissions: the female owner’s iPhone 16 Pro and the male owner’s iPhone 15 Pro Max, with corresponding device model identifiers of iPhone 17.1 and iPhone 16.2, respectively.

The vehicle’s backend data shows that during the time window described by the user, the vehicle received a parking assist command from the iPhone 15 Pro Max, which activated the parking assist feature (activation requires the vehicle to be within close range of the phone’s Bluetooth connection) and initiated the exit.

When contacting online customer service regarding the user’s feedback, they stated that the parking assist command originated from an iPhone 16. We have verified that our online customer service representative confused the device model identifier (iPhone 16,2) with the corresponding device model (iPhone 15 Pro Max) during communication with the user, leading to misunderstanding and miscommunication. We apologize for this and will continue to improve our service efforts.

Xiaomi Auto always prioritizes user safety and user experience. Thank you for your attention and support.

Owner’s Response:

I am the owner of the vehicle in question. The above information is true. The Xiaomi Auto team has verified the data with me in person and confirmed it is correct. Thank you for your professionalism and service. Please refrain from spreading rumors. Thank you.

Sources: Weibo, Xiaomi

In August of 2019, a man driving his Ford Explorer Sport Trac was hit from behind by another driver in a 2019 Tesla Model 3 who was using Autopilot. The Ford’s driver lost control as the truck rolled and his 15-year-old son, who was not wearing a seatbelt, was ejected and killed. Just weeks before the case was set to go to trial, Tesla quietly settled the lawsuit that followed.

The Tesla driver, Romeo Yalung, was traveling at 69 mph (111 km/h) in lane three of Interstate 880 in California. Video from the EV shows the driver of the Ford, Benjamin Escudero, signaling and moving into the same lane ahead. Neither Yalung nor the Tesla, which had Autopilot engaged at the time, slowed down to avoid hitting the Ford.

More: Crash Victim Trapped In Burning Tesla Sues Over Door Handles

The lawsuit named multiple parties, not only Tesla and Yalung but also Yalung’s wife, who was in the passenger seat at the time. Over four years, Judge Rebekah Everson had the option to dismiss the case but allowed it to move forward. Now, with trial just a month away, Tesla has reached a settlement for an undisclosed amount.

That might seem like a shock considering some of the surrounding factors. Tesla has a largely victorious record in court when lawsuits about Autopilot or Full Self-Driving (Supervised) come up. Juries and judges have agreed in almost all cases that the person behind the wheel is responsible for driving and not Tesla. In addition, video of the tragic crash certainly calls into question Mr. Yalung’s attentiveness.

That said, this all comes after a Florida jury decided that Tesla was 33 percent liable for another fatal crash involving Autopilot. In that case, the driver of the Tesla hit two people on the side of the road, admitted fault to the police, and said that he wasn’t paying attention when the crash happened. Despite that, the jury awarded the plaintiffs $242 million.

Tesla is now appealing that verdict, but another loss in court could have been damaging. Settling this California case may have been a calculated move to avoid the risk of a repeat outcome.

Sources: Freightwaves, Carcomplaints

As cars edge closer to taking over more driving duties, the question of how they interact with the people around them becomes increasingly important. Technology may be ready to let drivers take their hands and eyes off the task, but it also raises a new challenge: how does a car signal its intentions to those outside it? In that spirit, Opel has built a Grandland prototype with a lighting system designed to speak directly to pedestrians.

More: Skoda’s Illuminated Grille Shows Pedestrians When It’s Safe To Cross The Road

The test model is equipped with Level 3 driver assistance, which allows the person behind the wheel to disengage under certain conditions. Once the vehicle is in charge, though, a fresh layer of communication is needed to make sure people on the street understand what the car is about to do.

Lights As A Language

The Grandland prototype has been fitted with a new camera system that can detect objects and gestures, combined with an AI-backed “intention prediction” function. This allows it to better recognize real world scenarios like when a pedestrian emerges from parked vehicles or a child steps on the road when chasing a ball.

When operating in Level 3 autonomy (or higher), the SUV’s front and rear indicators shine in cyan. This color was chosen as it is not associated with any other vehicle function. When a pedestrian is detected in the path of the vehicle, it starts braking and the LED headlights change from white to magenta.

Furthermore, a small screen that normally displays Opel’s Blitz emblem shows a warning sign or a pedestrian alongside a halt gesture. When the car has come to standstill, the LEDs turns green and the display shows a walking figure. In more complex situations when the AI algorithm can’t figure out what to do, the driver is prompted to take control of the vehicle.

Rethinking Headlights

Philipp Röckl, Global Lead Lighting at Stellantis, explains the thinking behind the project:

“Headlights clearly play an important part in increasing road safety – but modern lighting systems are capable of so much more. For example, lighting signatures can easily be animated and used for animations. And the illuminated Blitz on the Grandland can be substituted by a display that can both show the illuminated emblem and be used to communicate with other road users and pedestrians. By using existing lighting elements, it makes it easy to add communication features once we will introduce autonomous driving features in our cars”

The Grandland with the unique equipment and lighting features follows a number of prototypes and production vehicles from other brands that examine the same issue.

The model was created in-house in Opel’s Russelsheim headquarters in Germany and was showcased at the International Symposium on Automotive Lighting (ISAL) in Darmstadt. It is the result of a collaboration between the Opel Concept Workshop and students from the TU Darmstadt university.

More: The Stellantis Hot Hatch You’ll Never Buy But Can Still Drive At Home

Stellantis and TU Darmstadt have formed a strategic partnership for joint research into new lighting technologies since 2022. As part of this, the automaker is currently funding three doctoral students at the university’s Department of Electrical Engineering and Information Technology.

On her part, Julisa Le, Stellantis’ Lead Innovation Engineer, said:

“We specifically chose colors that are not already associated with other vehicle functions to avoid misunderstandings. For example, red is intuitively understandable, but it is already used for stop lights or the rear lights of cars. Cyan and magenta have been thoroughly evaluated for perception purposes and are currently not present in any traffic situations. Both ensure clear, unambiguous communication with other road users.”

Opel

More than 115,000 examples of the hugely popular Xiaomi SU7 are being recalled in China due concerns with its driver-assistance systems. The move follows a high-profile crash about six months ago in which three college students were killed when their SU7 struck a concrete divider at 97 km/h (60 mph) while its Navigate on Autopilot system was active.

More: First Fatal Xiaomi SU7 Crash Sparks Questions About Self-Driving Tech And Locked Doors

China’s State Administration for Market Regulation has revealed that the driver-assistance system used by the SU7 has “insufficient recognition capability and may not adequately detect and warn drivers in certain scenarios,” as required by Level 2 or Level 2+ systems. It went on to specify that if the highway pilot assisted driving feature is enabled, it may not provide adequate warnings of performance in extreme scenarios.

Scale of the Recall

A total of 116,887 examples of the SU7 are being recalled, all of which were manufactured between February 2024 and August 2025. This accounts for roughly 38 percent of the 305,055 SU7s that had been sold across China as of July since the car’s early 2024 launch.

Xiaomi has confirmed that the fault will be corrected via an over-the-air software update, avoiding the need for physical servicing.

This recall coincides with the rollout of new safety standards in China for cars equipped with Level 2 driver-assistance systems. These regulations, which set stricter performance requirements, will come into effect in 2027.

Fatal March Crash

The most serious incident linked to the SU7 occurred in March on the Dezhou-Shangrao Expressway. Chinese media reported that at the time, the car was allegedly traveling at 116 km/h (72 mph) with the Navigate on Autopilot engaged.

The system was said to have flagged an “obstacle ahead” and initiated braking while approaching a construction zone. The driver intervened, steering left and applying moderate braking, then gradually turning back to the right. Moments later, the vehicle collided with a concrete divider and caught fire. All three occupants lost their lives.

Read: Chinese Smartphone Giant’s European EV Push Might Spell Trouble For Tesla

“A recall due to a faulty driver-assistance system may deter some Chinese consumers from buying Xiaomi cars,” said Chen Jinzhu, CEO of consultancy Shanghai Mingliang Auto Service, in comments to the South China Morning Post. “But the recall also shows Xiaomi’s commitment and confidence in resolving the issues.”

In addition to building some of China’s most advanced electric cars, Xpeng has been quietly pursuing something far more ambitious: an electric vertical take-off and landing (eVTOL) aircraft. The project has been in development for over a decade, and the company is even constructing a purpose-built factory to produce them.

That long-term vision, however, hit turbulence this week when two of its aircraft reportedly collided and crashed at the Changchun Air Show.

The two aircraft involved in the incident are understood to be the same models designed to fold up and slot into the rear of the six-wheeled Land Aircraft Carrier. Each uses a six-rotor layout with propellers and arms that can fold away, and they are built to operate either under manual control or in autonomous flight mode.

Folding Wings and Six Rotors

It remains unclear whether the test flight was being conducted under manual control or via the autonomous system. What is known is that the accident happened during a rehearsal ahead of the public show’s opening on Friday. One passenger is reported to have been injured, though not seriously, according to Guancha.

Read: Xpeng’s $280,000 AeroHT Mothership Launches 2-Seater Drone From Its Boot

In a statement issued to CNEVPost, Xpeng said that the two aircraft were participating in a formation drill but collided after flying too close to each other. Fortunately, one of the aircraft landed safely, but the other sustained more significant damage and caught fire after it fell from the sky.

Ambitious Production Goals

Despite the setback, Xpeng wants to start production of the six-wheeled AeroHT ‘mothership’ and the eVTOL for paying customers by 2026. Prices will start at approximately 2 million yuan or about $281,000 at current exchange rates, and it’s said that the company has already received thousands of orders for the pricey recreational vehicles. The eVTOL can reportedly operate at up to 500 meters.

Both vehicles will be built at a 180,000-square-meter site in Guangzhou’s Development Zone. The new factory is designed to handle up to 10,000 units per year, underscoring the scale of Xpeng’s ambitions in taking mobility from the road to the sky.

Tesla has fought, and consistently won, court battles over its semi-autonomous Autopilot and Full Self-Driving (Supervised). That record changed recently when a Florida jury decided that the EV maker was partially to blame for a crash that ended in a fatality. Now, Elon Musk’s company is pushing for a new trial, arguing the verdict could stifle development of safety technologies.

A Fatal Florida Crash

The case stems from the death of 22-year-old Naibel Benavides Leon, who was killed in 2019 when a 2019 Model S slammed into a parked Chevrolet Tahoe in Miami-Dade county. Her boyfriend, Dillon Angulo, suffered serious injuries.

More: Over 10,000 Owners Sue Tesla Over This Widespread Complaint

The driver, George McGee, admitted he dropped his phone, took his eyes off the road, and believed the car would brake on its own. At the same time, he conceded that he was negligent and placed too much trust in the car.

Jury Finds Tesla Partly Liable

Despite that, a jury found Tesla 33 percent liable, awarding $42.5 million in compensatory damages and a staggering $200 million in punitive damages. Jurors were reportedly swayed by the plaintiffs’ argument that Tesla overstated the capabilities of Autopilot, even as the company repeatedly warned drivers to stay alert and keep their hands on the wheel. According to CarComplaints, Tesla admits that the whole thing was a tragedy but says it was entirely the fault of McGee’s “extraordinary recklessness.”

Tesla Pushes Back

In its filing, the company said “No other car in existence would have stopped when the driver was telling it to ‘go.’” Tesla argued that McGee was “reckless in the extreme by ignoring or overriding every safety feature in his car,” including by pressing the accelerator pedal, which overrides the system’s cruise control and braking functions. There’s no question that this is a sticky situation, but Tesla makes a few interesting points.

“For as long as there have been cars, there have been reckless, self-absorbed drivers like McGee,” Tesla said. “Those drivers should face every legal consequence for their wrongful conduct. Holding Tesla liable for providing drivers with advanced safety features just because a reckless driver overrode them cannot be reconciled with Florida law. That rule would impede the development of safety features, deter progress, and cost lives both now and in the long run.”

The Bigger Question

Ultimately, all of this seems to stem from the nomenclature and advertising of Autopilot. Were it called something else and marketed differently, it would seem more difficult to blame Tesla at all. The automaker is asking the judge to either order a new trial or reduce the damages, a move that will weigh driver responsibility against driver-assistance technology. The outcome could make a huge impact in the future of automakers and their relationship with autonomy.

Toyota has launched a curious new product aimed at “various mobility services.” The Toyota e-Palette is a fully electric shuttle designed to work with autonomous driving hardware and software, and it can double as a mobile shop or service hub. After years of concepts and prototypes, the quirky EV has officially gone on sale in Japan, though its asking price may drop a few jaws.

From Concept To Reality

The project began with the e-Palette concept, which made its debut at CES in 2018. It was followed by a redesigned version, created for the 2021 Tokyo Olympic Games. Since then, Toyota has continued refining the design, culminating in the first commercially available e-Palette variant.

More: Toyota Launches Most Rugged GR Sport Land Cruiser Yet

Visually, it retains much of the original concept’s DNA. Its symmetrical, boxy silhouette, tiny wheels, and large glass panels give it a distinctly shuttle-like character. The “friendly” face is defined by round LED headlights, customizable digital signage, and a repositioned Toyota badge sitting above the bumper intake.

For production, Toyota added larger side windows and slim indicators on the windshield frame, while the rear end mirrors the front aside from red taillights and solid pillars.

Toyota

Fits Up To 17 People Inside

Unlike earlier prototypes that only offered passenger seating, the production e-Palette adds a central driving position with a yoke-style steering wheel and a steer-by-wire system. The cockpit features a central digital instrument cluster, an infotainment display on the right, another touchscreen on the left, and a fourth monitor for cabin monitoring. Essential functions are handled by physical buttons on twin control panels.

More: Toyota’s New Minivan Lets You Create A Lounge, Office Or Bedroom Inside

A low, height-adjustable floor, wide sliding doors, and a deployable ramp make access straightforward, particularly for wheelchair users and parents with strollers. Inside, four fixed seats line the rear, complemented by three side-hinged folding seats in the center. Toyota says the cabin can hold up to 17 people in total, including 12 standing, four seated, and the driver. That’s less shuttle, more party bus.

Buyers can also opt for a wheelchair lock system and a large overhead digital display, echoing public transit design. More importantly, the interior layout can be customized, transforming the EV into an entertainment pod or a pop-up store. Toyota imagines it shuttling commuters in the morning, transforming into a food truck at lunch, then acting as a sports viewing pod in the evening while charging.

Autonomous-Ready Shuttle

While earlier prototypes boasted full autonomy (with an operator for safety), the production e-Palette is currently limited to Level 2 driver assistance. However, Toyota stresses it is “capable of supporting automated driving” through optional third-party hardware and software upgrades called “Automated Driving Kit”.

More: Waymo Quietly Moves Ahead Of Tesla In The Race For Robotaxis

The automaker plans to conduct automated driving demonstrations with dealers and local governments in Japan. Their goal is to launch vehicles with Level 4 autonomous systems in the fiscal year 2027, which ends on March 30, 2028.

Fully Electric With A 160-Mile Range

The e-Palette measures 4,950 mm (194.9 in) in length, roughly the same as a Land Cruiser Prado. Where it differs is in width and height, stretching to 2,080 mm (81.9 in) wide and 2,650 mm (104.3 in) tall, giving it a much boxier footprint than the truck-based SUV. It also weighs in at a hefty 2,950 kg (6,504 lbs).

Power comes from a single electric motor delivering 201 hp (150 kW / 204 PS) and 266 Nm (196.2 lb-ft) of torque, with top speed capped at 80 km/h (50 mph). That’s a big step up from the 20 km/h (12 mph) top speed of the Olympic Games prototype.

A 72.82 kWh battery provides up to 250 km (160 miles) of range. Toyota claims the battery can charge to 80% in around 40 minutes with rapid charging, or 12 hours from a standard outlet. Furthermore, the shuttle can also serve as an emergency power source.

An Extra-Terrestrial Price

All of this sounds appealing until you get to the price. The Toyota e-Palette is listed in Japan at an eye-watering ¥29 million ($196,400) before options. This is almost double the price of the GR Supra Final Edition, surpassing even the ¥27 million ($182,800) price tag of the Century SUV flagship, making it the most expensive model in Toyota’s range.

Generous subsidies from Japan’s Ministry of the Environment can cut as much as ¥15,835,000 ($107,200) off the sticker, softening the blow. Toyota hasn’t revealed production numbers, but notes the e-Palette will be built to order, which partly explains the sky-high cost.

Initially, the model will be used in the Toyota Arena Tokyo and the surrounding areas, as well as in the futuristic Toyota Woven City in Japan. Below, you can check out videos of earlier e-Palette prototypes.

Tesla might be in the midst of trying to expand its Robotaxi platform as fast as it can, but Waymo is bounding ahead. It just announced plans to bring its fleet to Denver, Colorado, and Seattle, Washington—though for now, the vehicles will still have human drivers behind the wheel as part of an initial testing phase. This step is designed to fine-tune operations and ensure the cars are ready for fully autonomous rides in the future.

Read: Waymo’s NYC Debut Comes With Training Wheels

Denverites can expect to see a mixed Waymo fleet of Jaguar I-PACE SUVs with its fifth-gen Waymo Driver and Zeekr RTs equipped with the sixth-gen system. Both vehicles will have human drivers behind the wheel to begin. They’ll gather data about the entire city while Waymo coordinates with state and city officials.

Local Voices Onboard

Speaking of those officials, they sound enthusiastic about the move. Governor Jared Polis said, “I’m excited to help Coloradans get where we want to go faster, safer, with additional transportation options.” Mayor Mike Johnston added, “Waymo’s innovative, climate-friendly technology will not only make our streets safer but cleaner, and I can’t wait for my first ride.”

At the same time, Waymo is doing something almost identical in Seattle. Human drivers will guide the cars there while collecting data about traffic patterns before handing over control to the fully autonomous cars at some stage in the future. Waymo emphasized that its years of experience in areas with heavy rain, snow, and complex urban layouts will help it make its service in both locations.

View this post on Instagram

A post shared by Waymo (@waymo)

Expanding The Map

This new Denver and Seattle test phase is all just a tiny piece of a much larger expansion going on for Waymo. The company already operates fully autonomous cars with no human supervision in Phoenix, Los Angeles, Austin, and Atlanta. It’s also working on expanding to Miami, Washington, D.C., New York City, and other locations.

In fact, it recently announced that the San José Mineta International Airport just gave it the green light to begin service on airport grounds. Tesla might have more coverage in Austin, Texas, but this proves that Waymo is still far ahead in the autonomous taxi business.

View this post on Instagram

A post shared by Waymo (@waymo)

Credit: Waymo

Love him or hate him, Elon Musk has built a reputation for sometimes making the impossible seem inevitable. From electric cars dominating global markets to rockets that land themselves with pinpoint precision, his record is extraordinary. Yet, even for someone with those successes, the 2025 CEO Performance Award just presented by Tesla’s board represents a challenge of unprecedented scale.

Read: Tesla Just Quietly Fixed A Problem Owners Have Complained About For Years

The plan could grant Musk over 423 million shares and a total compensation package of around $1 trillion. Of course, that only happens if Tesla hits some staggering milestones. The award is split into 12 tranches, each unlocking as Tesla hits specific market capitalization targets.

According to the SEC filing detailing the package, it goes as follows: the first tranche at $2 trillion, nine more at $500 billion increments, and the final two at $1 trillion each, culminating in a mind-bending $8.5 trillion valuation by 2035. Right now, Tesla is worth a little over $1 trillion.

A Trillion-Dollar Tightrope

Market cap is only a small piece of the story, though. Musk must also achieve at least $400 billion in sustained annual adjusted EBITDA and hit operational targets that push Tesla into uncharted waters. Among them: putting 1 million fully autonomous Robotaxis into commercial service and delivering 1 million Optimus AI bots.

Market cap is only a small piece of the story. Musk must also deliver at least $400 billion in sustained annual adjusted EBITDA, which is finance-speak for the company’s earnings from its core business before accounting for things like taxes, interest, or non-cash costs such as depreciation.

On top of that, he faces operational goals that push the company into uncharted territory, such as putting 1 million fully autonomous Robotaxis into commercial service and delivering another 1 million Optimus AI robots.

Today, Robotaxi still requires a safety driver in the car, so these targets aren’t theoretical; they demand breakthroughs in autonomy, scaling, and execution. It’s a package that forces a hard look at Musk’s history. He’s undeniably a visionary, but he’s often over-promised and under-delivered. He’s often been late, sometimes very late, about tech promises. That’s to say nothing of his political engagement.

Betting His Legacy On Autonomy

At the same time, if he can actually achieve everything laid out in this compensation package, it’ll be hard to argue that he’s not the defining innovator of our era. That said, the board has set up this award in a way that Musk can’t simply hit the goals and cash out.

He must remain at Tesla for the next 7.5 to 10 years. The goals must be met by 2035. Tesla said it “believes that Mr Musk’s vision and leadership are critical to nailing that execution”.

At this point, all of Musk’s promises of autonomy are all the more intriguing. His salary and perhaps his legacy rely on him being right about Tesla leading the autonomy charge sooner rather than later.

Tesla’s Full Self-Driving (Supervised) system, along with its ambitious claims, has repeatedly drawn the company into controversy, and it now faces yet another round. The company is once again facing legal trouble, this time after a U.S. District Judge in California ruled that Tesla must answer a certified class action alleging it misled drivers about the self-driving abilities of its vehicles. Tesla had argued the case should be dismissed, but the court disagreed.

Read: Musk’s Robotaxi Pitch Just Backfired And Shareholders Are Suing

The automaker has consistently promoted the idea that all vehicles it built since 2016 came equipped with hardware capable of full self-driving, albeit under supervision. These assurances were made across Tesla’s website, blog posts, social media channels, and directly by chief executive Elon Musk.

In practice, though, the cars have not lived up to those promises. Tesla also asserted that vehicles with its Full Self-Driving package would eventually deliver Level 4 and Level 5 autonomy, but neither has materialized.

Judge’s Assessment

U.S. District Judge Rita Lin noted that claims about Tesla vehicles lacking the necessary hardware for autonomous driving, combined with the company’s failure to “demonstrate a long-distance autonomous drive with any of its vehicles,” provide grounds for lawsuits brought by two groups of drivers.

Tesla does not engage in typical mass advertising, and the Judge noted that ordinarily, the channels it used to promote its self-driving hardware and software may not be “enough to establish a class-wide exposure for a traditional car manufacturer.”

However, she said it’s reasonable to infer that class members went to Tesla’s website for information on its Full Self-Driving (Supervised) technology. She added that thousands of people likely saw a claim on Tesla’s website from October 2016 to August 2024 that said its vehicles contained the hardware necessary for fully autonomous driving.

Tesla’s Defense

Tesla countered that it is unreasonable to assume all class members saw those statements. The automaker also argued there is no unified proof showing the claims were significant enough to influence purchasing decisions, according to Reuters.

The class actions in California include drivers who purchased the Full Self-Driving Package from May 19, 2017, to July 31, 2024, and who opted out of Tesla’s arbitration agreement, as well as drivers who purchased the package from October 20, 2016, to May 19, 2017.

In the US, Tesla’s arbitration clause requires all disputes to be resolved through arbitration rather than in court, unless a purchaser or lessee opts out of the clause within 30 days of buying or leasing a Tesla vehicle.