AI in the Automotive Industry

There isn’t an automotive brand today without a blueprint for applying artificial intelligence to their product line. The scope ranges across vehicle design and engineering, manufacturing and quality, and customer experience and in-car technology. The overall trajectory of AI in the automotive industry is clear: the technology is rapidly moving out of the experimentation phase.

How much of this is consumer demand-driven, and how much is an industry caught up in trying to win market share with a new technology? Automakers are clearly under pressure to compete on software and AI capabilities. Still, it is natural to assume customers care less about technology for technology’s sake than about what it does for them. Understanding that difference will become increasingly important as automotive companies decide where to invest.

How are automotive companies applying AI?

A review of several major manufacturers can help ground us in the different AI strategies that are emerging.

General Motors: data from the road

In April 2026, GM announced that with 23 Super Cruise-equipped production models, it had accumulated over a billion hands-free miles from 750k vehicles. The company celebrated the milestone, but what is more important than miles driven is what it is doing with the data generated from those miles. According to the company, its real-world driving data not only improves current Super Cruise abilities but is also being used to train AI models and inform the development and validation of future autonomous systems.

We don’t have to wait long for the results of this strategy, as GM has already begun road testing its next-generation automated assistance technology. Further, the company plans to introduce eyes-off driving beginning in 2028 with the Cadillac Escalade IQ. In effect, the company has turned a fleet of deployed vehicles into a source of continuous information for future product development, with each subsequent version reinforcing the learning from the last. This marks a significant departure from the traditional model, where testing happened before a vehicle entered production, and it is a strategy we are likely to see across the market as a whole.

Hyundai: data in a flywheel

Hyundai Motor is adopting a strategy that takes the idea further. The company is deploying a “data flywheel” approach to improve its Software-Defined Vehicle (SDV) line. In this model, data collection and analysis, AI and service enhancement, and over-the-air (OTA) updates will form a virtuous circle—where the positive benefits of each element drive the next in a compounding loop of continuous improvement.

The most noticeable effect for consumers will begin with the Grandeur, when the company introduces Gleo AI, its generative AI agent. As customer experiences with the AI agent grow, so too will the company’s understanding of its customers. It plans to use the same approach for autonomous vehicles (AVs) as well. The company expects customer data to grow so quickly that it plans to launch a 100-megawatt data center with more than 50,000 GPUs in 2029. The data center will unify data gathered from a growing fleet of AI-enabled vehicles with its integrated in-house AI, forming the foundation for the company to compete in both SDVs and AVs in the future.

Again, the significance isn’t that generative AI is coming to Hyundai dashboards. It is that the company is implementing a strategy in which its vehicles generate the data used to shape future software, features, and vehicle development.

What is working, and what is not?

Consumer acceptance and demand are less straightforward. According to AAA, only 13% of U.S. drivers consider self-driving development a priority, down from 18% in 2022. And 60% report they are still afraid to ride in a self-driving vehicle. Given the industry’s push to develop and deploy AVs, this is a troubling finding.

While consumers are apparently apprehensive about AVs, they are, on the other hand, interested in features that enhance safety. This contradiction illustrates the potential divide between the promised safety of a future inhabited by autonomous transportation systems and the safety technologies drivers already understand, trust, and see working in the vehicles they drive today.

At the same time, deployment and use of assistive technology are moving faster than consumer understanding. An IIHS study of regular users of GM’s Super Cruise, Tesla’s Autopilot, and Nissan’s ProPILOT Assist found that many respondents were more likely to take their hands off the wheel and eyes off the road to perform non-driving-related activities. Even more worrying, drivers reported that factory safeguards for assisted-driving features were annoying and that they actively worked to bypass them.

These findings suggest that new technological capabilities in vehicles will not, in and of themselves, ensure adoption. Rather, manufacturers will also need to build trust and clearly communicate what these systems can and cannot do.

What does this mean for automotive market research?

As AI in the automotive industry becomes more widespread and part of the fabric of transportation, manufacturers still have to answer a much more important question: what will drivers actually value?

As vehicles become more software-defined and increasingly capable of learning from their users, market research will also need to evolve. Automotive market research, including car clinics, focus groups, ethnographic research, concept testing, and customer surveys, can help brands understand which AI-enabled features solve real problems or meet consumers’ true needs.

The challenge is not simply to determine whether consumers “want AI.” It is to identify the capabilities they understand, trust, and find useful enough to pay for. For business leaders making long-term bets on autonomous driving, intelligent cockpits, and software-defined vehicles, that distinction can determine which investments create value and which become expensive technology customers never asked for.