Last year, I got a real world taste of how artificial intelligence is gearing up to change the way we interact with our cars. While driving the new Audi Q6 e-tron, I asked, “Hey Audi, what’s a good place to visit outside of Bilbao?” and was pointed to the beautiful coastal city of San Sebastián and told to check out the stunning views from Monte Igueldo or the Peine del Viento sculptures by Eduardo Chillida.
The car I’d tested the week prior could barely “find fast food along my route,” but thanks to integration with ChatGPT, the electric Audi was able to understand the full context of the question, find the answer on the web and help me plan a day trip with normal human-sounding sentence. That’s just the tip of the automotive AI iceberg.
In 2025, AI is poised to play a significant role in shaping the future of transportation, promising to revolutionize the way we design, manufacture and interact with cars, trucks and SUVs. That means enhanced safety, convenience and personalization are on the horizon. AI dashboard technology has the potential to minimize distractions and mitigate the consequences of lapses in attention, particularly as driver assistance and autonomous technologies continue to evolve.
There’s also the rise of driverless robotaxis, which are expanding operation to more cities in 2025. However, as these technologies become more prevalent in vehicles, it falls to automakers and regulators to prioritize safety and security within these AI-powered vehicles.
Future trends in AI-powered automotive technology
AI in cars isn’t exactly new. During my decade and a half as CNET’s car tech expert, I’ve seen automakers and startups using artificial intelligence to design and power everything from advanced driver assistance features and natural language voice control to the way vehicles are shaped aerodynamically and safely deform in accidents. However, the recent AI gold rush has spurred increased attention and investment in robotics and large language model technology — along with the rise in prominence of large dashboard screens, always-on, high-speed connectivity and over-the-air updatable software — has accelerated AI’s effect and visibility in the most high-tech production vehicles.
In 2024, we saw a myriad of automakers and startups announcing AI technologies large and small hitting the road “soon.” Here’s what you can expect in 2025 and beyond.
Autonomous vehicles are here… sort of
The self-driving car is no longer a distant dream. Today you can open the Waymo One app in San Francisco, Los Angeles or Phoenix and be chauffeured to your destination in a driverless Jaguar i-Pace adorned with a full suite of radar, lidar and optical sensors. The service isn’t perfect — crashes are rare, but Waymo cabs can be brought to an annoying halt by confusing circumstances and there was the weird honking incident — but, with the aid of remote operators to get passengers out of the stickiest situations, Waymo One has proven fairly reliable. In 2025, Waymo expects to expand service to more cities, including Atlanta and Austin in partnership with Uber.
Motional, a Hyundai x Aptiv joint-venture startup, has been testing a robotaxi of its own for the last year. In 2025 it’ll be using the $475 million investment it’s receiving from Hyundai to push development and expand operation of its Hyundai Ioniq 5-based robotaxi, which is currently in service in Las Vegas and Los Angeles in partnership with Uber and testing in cities like San Diego, Boston and Pittsburgh.
Enlarge Image
Tesla’s new robotaxi is estimated to arrive in 2026 without a steering wheel or pedals.
Robotaxis are cool and all, but what if you want to own your own self-driving car? Last year Tesla announced its long-awaited Cybercab, which is expected to cost “under $30,000” and arrive sans a steering wheel or pedals sometime in 2026. Tesla head Elon Musk claims that owners will be able to summon the scissor-doored coupe via an app, choose their destination via touchscreen and arrive at their destination auto-magically.
Tesla’s reliance fully on optical sensor tech — no spinning lidar humps here — has me concerned over safety, but Tesla touts the advantage of eventually making the tech available on current Model 3 and Model Y vehicles equipped with its Full Self Driving tech via OTA update. Tesla is no stranger to delays: Musk has been promising “fully autonomous robotaxis” coming “next year” since 2019, so who knows when it’ll actually arrive. That said, Musk’s newfound political influence may help smooth the way.
Watch this: Tesla Cybercab: A Robotaxi of Your Own for Under $30k
04:32
We’ve cooked up an even deeper dive into what you can expect from robotaxis and autonomous cab services in 2025, which you can check out here.
Smart driver assistance systems are getting smarter
A vehicle needn’t be fully autonomous to be enhanced by the capabilities of AI, which is also making advanced driver assistance technologies safer and more convenient. Features like adaptive cruise control, lane departure warning and automatic emergency braking are combining and becoming increasingly sophisticated, thanks to AI-powered algorithms.
General Motors’ Super Cruise technology allows drivers to go hands-off on about 750,000 miles of approved highways with automated steering, acceleration and braking, requiring the driver to only keep their eyes and attention on the road for safety. The tech has been around since 2017 but has seen many enhancements and improvements over the years, including the ability to tow hands-free in the upcoming GMC Sierra Denali EV. In 2025, Super Cruise will be available in around 15% of General Motors’ lineup, around 22 vehicles in total including the upcoming Cadillac Optiq EV.
Enlarge Image
I’d be shocked if the upcoming CLA-Class doesn’t feature Mercedes-Benz’s groundbreaking Level 3 Drive Pilot system.
In California and Nevada, drivers of the Mercedes-Benz EQS sedan and S-Class models recently gained access to the automaker’s Drive Pilot system, the first Level 3 autonomous system approved for use in a production car on public roads. Drive Pilot allows drivers to go hands-free under certain conditions without the requirement to keep their eyes on the road, freeing up their attention to pass the miles doing other things. The automaker hasn’t announced specific plans to bring the tech to other vehicles, but I’d be shocked if the new CLA — featured at CES 2024 — doesn’t ship Level 3-ready.
Hal 9000 for the highway
Both Volkswagen and Mercedes-Benz made big splashes this time last year at CES with announcements that ChatGPT-powered voice assistance is coming to their vehicles’ dashboards. ChatGPT (and similar LLM-based technologies) promise more personalized, intuitive and natural in-vehicle experiences, including the ability to simply talk to your car like a human and receive knowledge-based responses. So, instead of the wooden “NAVIGATION… DESTINATIONS… COFFEE SHOPS… OPTION 3” with laborious pauses between prompts, drivers may eventually be able to say something like, “Hey car, take me to that pizza place Jon recommended in North Beach” and get on their way with much less fuss and distraction.
More than a few automakers (from Volvo to Ford to BMW and Mercedes) are also using AI technology to provide predictive and contextual information. So if your car learns that you usually pick your kids up from school in the afternoon, it may automatically begin suggesting that when you buckle up at 3 p.m., or if you’re a hybrid worker with a long commute into the office twice a week, the car may be able to adjust your charging schedule automatically.
Enlarge Image
The Audi Q6 e-tron’s AI-powered voice assistant is one of the best and most natural I’ve ever used.
The aforementioned upcoming Q6 e-tron electric SUV, for example, boasts one of the most impressive and natural voice assistants I’ve ever used. I was recently able to test Audi’s latest-generation AI assistant in real world conditions and found that it understood a wide range of naturally spoken commands and questions and integrated with nearly every function of the vehicle’s operation beyond the normal navigation, messaging, audio and climate. For example, I was able to just tell the car that I wanted to turn off the audible speed limit warnings when they got annoying and be brought to the exact place in the menu system to make the change. When I stumbled upon a twisty road segment, I could just say, “Activate Dynamic Mode.”
And I could take advantage of ChatGPT integration to ask the Audi knowledge based questions like, “What’s a good place to visit outside of Bilbao?” — a simple enough question for a human, but it was impressive seeing the system understand that I was looking for tourist destinations within driving distance of the Basque city and return a simple, spoken answer with clear information about why I’d want to visit rather than a bulleted list on the screen.
The self-maintaining car
It seems a small thing, but integrating AI deeply into vehicle systems enables much more than just enabling you to ask the car questions about itself (“How much air should I put in the tires?”). AI can also be used to analyze data from a car’s sensors and onboard systems to predict vehicle maintenance needs, identifying potential issues before they become, well, issues. Software may be able to detect, for example, irregularities in performance, charging behavior or thermal management that the driver may not (or may not be able to) notice on their own, and use that information to warn the owner to have the vehicle serviced before a more expensive problem develops, reducing the time a vehicle spends in the shop waiting for big fixes and improving overall reliability.
AI can also help your vehicle communicate what is wrong with itself in a way that’s more easily understood by a human. For example, natural language can be used to explain that your check engine light is a fix you can handle yourself — say, a loose gas cap — or a bigger problem that requires more urgent service. Perhaps one day, even, your car will know based on your calendar that it’s going to be parked for an extended period because of a vacation on your calendar and can take steps to condition its own battery for longer term storage. AI smarts could eventually lead to cars that basically maintain themselves.
The factory of the future
At the tip of the automotive spear, AI is transforming the way automakers design and manufacture cars, improving efficiency on the road and on the factory floor. From using AI-powered simulation to shape the aerodynamics, crashworthiness and assembly methods of future cars still on the drawing board to refinements in production to supply chain management, AI is being used to optimize every aspect of vehicle manufacturing.
Automakers are even using AI to design the factories themselves, simulating the entire manufacturing process virtually before reshaping and reorganising how the cars move from sheet metal to assembly to customization and paint. These AI-powered factories — along with new modular vehicle construction — are what allows BMW to build the fully electric i4 sedan on the same assembly in its Munich plant as the hybrid and plug-in hybrid cars like the 3 Series. This allows BMW to adapt quickly to demand for all of its powertrain options as policies and market conditions change. (Though, Bimmer has announced plans to switch the Munich plant exclusively manufacturing its next-generation Neue Klasse EVs in 2027.)
Enlarge Image
Hyundai’s Singapore Innovation Center is a robotic factory without a fixed assembly line.
At the cutting edge of AI-powered manufacturing is the Hyundai Motor Group’s Singapore Innovation Center, a factory without fixed assembly lines. I visited Hyundai’s AI-powered factory where consumer and robotaxi versions of the Ioniq 5 are being built and was wowed by the high level of automation and the way vehicles moved freely around the modular factory floor on robotic sled, stopping at stations as needed to receive parts or bypassing whole parts of the assembly line not required by a specific trim. Along the way, AI-powered Boston Dynamics Spot robotic dogs helped human workers with inspection and quality checks, while a complex software allowed just a handful of people to oversee the entire production.
Last year, Hyundai started cranking electric SUVs out of its new $7.6 billion Metaplant America outside of Savannah, Georgia (my favorite city in the world), starting with its award-winning Ioniq 5 EV. In 2025, the new, much larger Ioniq 9 will also begin production at the facility, and the automaker is also considering adding hybrid and combustion-powered SUVs, a mix that could only be tackled efficiently with the AI-powered lessons learned in Singapore.
The big impact of AI in the automotive industry
The widespread adoption of AI in the automotive industry is expected to have a profound impact on consumers, manufacturers, suppliers and regulators.
Consumers: Consumers can expect safer, more convenient, and more personalized driving experiences. Hopefully, AI-powered dashboard tech can help limit distraction, while smarter driver aid and autonomous tech will eventually mitigate the consequences of lapses in attention.Manufacturers and suppliers: Manufacturers can benefit from increased efficiency, improved quality and reduced costs. How and how much of that cost savings will be passed on to the consumer is the subject of debate, but I believe that AI-powered manufacturing will help automakers adapt to changing demand and help minimize the sort of supply chain issues that caused new car prices to spike during the COVID-19 pandemic. Regulators: Meanwhile, regulators and governing bodies face the challenge of developing new frameworks to govern the use of AI in vehicles. What autonomous tech will be allowed and where? How can we ensure drivers use AI responsibly and that there are guardrails that prevent companies from abusing consumer data? These are questions that will be argued in the halls of government over the next few years. Elon Musk’s newfound political influence could result in stricter AI safety regulations, but his involvement in AI (with xAI) and automotive (Tesla) industries warrants scrutiny regarding the fair application of these standards.
What’s the catch?
While the potential benefits of AI in automotive are significant, there are also challenges that need to be addressed.
Safety and security: As more tech makes its way into the car, it becomes ever more important that automakers can ensure the safety and security of AI-powered vehicles. Whether we’re talking about how an autonomous car moves in a chaotic environment or how AI responds to driver inputs, voice commands or sensor data, robust testing and validation processes are essential to mitigate risks. AI hallucinations are kind of hilarious when you’re generating cat pictures, but a car that hallucinates at 80 mph sounds absolutely terrible. Ethical and legal considerations: From the oft-invoked trolley problem when considering how vehicles react to situations where there’s no clear “good” option; to questions of liability in the event of an accident; to discussions around how companies use and share consumer information collected by AI systems (will your car rat you out to your insurance company?); many of the questions surrounding AI and autonomy can be extremely nuanced and difficult to wrap your head around.Data privacy: The collection and use of data by AI systems in vehicles raise concerns about privacy. That includes information users may give AI access to — for example, granting access to email or a calendar to enable predictive features — and information deduced automatically by an AI system, such as when a car learns your scheduled habits and makes decisions based on that collected info. Whether you give it or not, AI-powered vehicles will know a lot about you and your habits and how that info is handled, stored and disposed of will be paramount.
Despite these challenges, the opportunities presented by AI in automotive are immense. As AI technology continues to advance, we can expect that we’ve only scratched the surface of what applications automakers have planned (and have yet to dream up) in the years to come.
