Automotive AI Roundup: Robotaxis, FTC Rules and Dealer Service Tech

A wider look at the week's automotive AI news beyond our feature on Tekion's new agentic suite: two more robotaxi bets, a Senate push on self-driving safety data, the FTC pricing crackdown's latest fallout, and a service-routing play that turns mobile repair into a capacity strategy.

---

Mobileye becomes an operator, not just a supplier

Mobileye, known for years as the company selling automated-driving technology to other automakers and fleets, is now planning to run its own robotaxis. The Israeli firm intends to launch a service by 2027, starting with roughly 100 vehicles in a single major U.S. city and scaling toward about 17,000 vehicles within five years of launch. That puts it directly against Waymo, Zoox, and Tesla in a field that has gone from pilot projects to paid rides in a handful of cities.

The interesting part is the role change. A supplier deciding to operate its own fleet means it now competes, at least partly, with the customers it sells to. For the broader market in automated cars, it adds another well-funded contender and another approach to the same problem of moving people without a driver. Mobileye's pitch has always leaned on a camera-and-radar perception stack it can produce at automotive scale, and an operator business gives it a way to capture more of the value its vehicle automation creates rather than selling components into someone else's robotaxi.

For dealers, robotaxis remain a long-horizon story rather than a this-quarter one. More automated car companies chasing ride-hailing does not change what happens on a showroom floor next month. But it does shape the decade: every robotaxi mile is a mile a household might not buy a car to drive, and the used-vehicle supply, service patterns, and ownership models that dealerships depend on all sit downstream of how fast this scales. A supplier turning operator is a signal that the people closest to the technology think the economics are finally close enough to chase directly.

Source: Automotive News

---

Stellantis, Uber and Wayve aim for Level 4 at global scale

Stellantis is teaming with Uber and the autonomous-driving software company Wayve to deploy Level 4 robotaxis worldwide, a partnership that lines up three different strengths behind the same goal. Stellantis brings the vehicles and manufacturing scale of a global automaker, Uber brings a ride-hailing network and demand, and Wayve brings the self-driving software that has to make the whole thing work without a person at the wheel.

The structure is the story. Where Tesla has tried to build the car, the software, and the service itself, this alliance splits the work among specialists. That partnership model has become the dominant pattern for getting artificial intelligence and cars onto public roads, because no single company has to be best at everything. An automaker that struggles to write world-class self-driving software can partner for it; an AV software firm that cannot build cars at volume can plug into one that can.

Level 4 means the vehicle can drive itself within a defined area without human intervention, which is the threshold a true robotaxi has to clear. Reaching it at "global" scale is a high bar, and announcements like this one tend to run ahead of deployments. Still, a major OEM committing to L4 through an alliance rather than a science project is a meaningful vote of confidence in vehicle automation as a near-term commercial business, not just a demo.

For the automotive retail world, the takeaway mirrors the Mobileye news: the companies that build and sell cars are increasingly betting that some future driving will not involve an owner at all. How dealers fit into a world with more automated cars and fewer driver-owned ones is the question every one of these partnerships quietly raises.

Source: Automotive News

---

Washington presses NHTSA over Tesla's self-driving safety numbers

U.S. senators, led by Edward Markey, sent a letter to National Highway Traffic Safety Administration Administrator Jonathan Morrison urging the agency to scrutinize how Tesla measures and promotes the safety of its Full Self-Driving system. The letter argues that Tesla has used misleading and incomplete statistics to claim FSD is far safer than human driving, and that the gaps expose weaknesses in how NHTSA collects autonomous-vehicle safety data in the first place. It draws on a recent Reuters investigation.

The specifics are worth understanding because they are about method, not just marketing. According to the letter, Tesla compared crashes in its own fleet that triggered airbags against a broader federal category that counts any vehicle towed from the scene, a comparison that can make FSD look safer than a like-for-like reading would. It also compared newer Teslas against the much older average vehicle on U.S. roads, and counted a crash as FSD-involved only when the system was active within five seconds before impact, versus the 30-second window NHTSA's own crash-reporting order uses. When a severe crash destroys a car's telematics, the senators note, no data gets reported at all.

This is the accountability question that follows every claim an artificial intelligence automotive system makes about itself. When the system that drives the car is also the source of the data used to judge it, regulators and buyers need a way to verify the numbers rather than take them on faith. The letter's deeper point is that NHTSA's framework does not yet require the exposure data, such as miles traveled, needed to compare safety honestly across companies.

For anyone deploying artificial intelligence and cars, dealers included as they adopt AI tools of their own, the lesson is that auditability is becoming the price of trust. Logs that show what an automated system actually did, and how that was measured, are turning into a regulatory expectation, not a nicety.

Source: U.S. Senate (Office of Sen. Edward Markey)

---

The FTC's pricing crackdown keeps reshaping how dealers advertise

Three months after the Federal Trade Commission sent identical warning letters to 97 dealership groups over advertised prices that omit mandatory fees, the pressure is still rippling through the industry. The original March letters cited six pricing practices the agency said could violate the FTC Act by showing consumers a number lower than what they actually pay. What has happened since shows how fast the ground is shifting.

In April, the FTC clarified that dealership social-media posts count as advertising and that competitors can report each other for noncompliant pricing. In May, TrueCar said it would fold mandatory dealer fees into its advertised prices, and the FTC publicly released the names of all 97 warned groups, including large chains like Lithia Motors, AutoNation, and Group 1 Automotive. Analytics firm Widewail found those warned dealers had drawn far more bait-and-switch complaints than the industry average. By June, S&P Global Mobility had opened a free fee-management tool, Cox Automotive had paused its price badges on listing sites, and more than 200 dealerships had registered for an industry compliance summit.

The AI connection is direct. When an ai auto sale chatbot or an automotive agent quotes a price to a shopper, the dealership owns that claim exactly as if a salesperson made it. As automatic car dealers lean harder on AI to answer questions and structure deals before a customer ever walks in, compliance has to extend to the tools, which means knowing what every automated quote said and being able to prove it. That is why auditability and human oversight, the same governance themes showing up across agentic AI, matter here too: a tool that quotes the wrong all-in price at scale is a liability multiplier.

As TrueCar chief executive Scott Painter put it, price transparency has been a founding principle, while industry voices like consultant David Spisak acknowledged dealers share culpability for the practices that drew regulators in.

Source: CBT News

---

Mobile service turns into a fixed-ops capacity strategy

While the headlines chase robotaxis, one of the quieter automotive workflow stories of the week is about getting a technician to the customer instead of the other way around. New York Honda and Acura dealer Brian Benstock is rolling out mobile service across three of his stores through software provider Curbee, which handles the routing and decides which repair orders can actually be completed off-site. Curbee estimates roughly 37 percent of repair orders qualify.

The adoption curve is steep: Curbee says about 300 dealerships had signed up as of January 2026, up from 25 a year earlier. Benstock frames mobile service less as a convenience perk and more as a way to free up capacity. Every 10 repair orders performed at a customer's driveway, he notes, free up 10 repair orders' worth of bays and lifts back at the store. His dealerships have picked up and delivered more than 200,000 vehicles since 2017, generating over $60 million in gross profit, so the economics of meeting customers where they are is something he has already tested at scale.

The technology underneath is the point. Deciding which jobs can leave the shop, sequencing them efficiently across a metro area, and matching them to the right technician is a classic routing-and-decisioning problem, exactly the kind of vehicle automation and software orchestration that pays off fastest in service, the most profitable part of many dealerships. Curbee chief executive Chandarana describes mobile service as "a capacity strategy," and that framing matters: the win is not a flashier customer app, it is more billable hours from the same fixed footprint.

For automatic car dealers weighing where to spend on AI and automation, fixed ops is an underrated answer. The customer-facing agents get the attention, but software that quietly squeezes more capacity out of the service drive often shows up faster on the bottom line.

Source: WardsAuto

---

Sources

• Mobileye to start robotaxi service in 2027, challenging Waymo, Zoox and Tesla, Automotive News
• Stellantis teams with Uber and Wayve to deploy Level 4 robotaxis worldwide, Automotive News
• U.S. Senate letter to NHTSA on Tesla Full Self-Driving safety statistics, Office of Sen. Edward Markey
• FTC's pricing crackdown: 3 months since 97 warning letters shook the industry, CBT News
• New York dealer boosts customer convenience with mobile service, WardsAuto

Related: Automotive AI Agents Graduate From Chatbot to Coworker