Just as Donald Trump has declared that North Korea must drop its nuclear program, lest they face an apocalyptic attack by the U.S., the auto industry must embrace the new digital realities, or face competition from competition that will not. The United States is by far the largest driver of vehicle sales, but China will overtake the U.S. in 2025, according to analysts at Navigant Research. As sales plateau in the developed countries, the auto industry will transform itself to drive future growth. Among new technology offerings, the most exciting and controversial is artificial intelligence. The advanced algorithms that can to match customer needs in increasingly complex programs are a massive opportunity for automakers.

Berard Marr, senior research analyst at IHS Markit, says that “Intelligent Models are still at the early stages of adoption” in the auto industry. Traditionally, if an auto company needed to change the way a vehicle operates, it would need to bring on board an outside third party–a “dumb” supplier–to help design and integrate the new system. But that system is now becoming smarter, more connected, and more open-source, which means that automakers can make the change themselves.

American companies’ economies of scale are unmatched, and as one might expect, they’re making the most of the advantages of scale–with plans to launch automated driving networks, use third-party data to do predictive analysis of roadside incidents (ranging from road traffic to pedestrians), and apply cognitive computing.

“In the past, when brands have done cognitive computing,” Marr says, “they’ve had to hire a firm that specializes in cognitive computing to do this, which has very high overhead and just takes time.”

He adds that AI use in the auto industry is the result of strong partnerships between companies such as Amazon, IBM, and Google. Most new vehicles are already equipped with cloud-based infotainment systems, and traffic cameras (often used to protect against distracted drivers), which are already making them smarter. But Marr says AI’s use will also be expanded to areas like “automated driving networks, autonomous zones, or traffic management.”

Another area where AI will be used is connected car services, or “infotainment systems,” as well as newer technologies like mapping and traffic safety. There are already projects underway in Ohio to expand VW’s map-based service by adding AI, road hazards, video, and real-time traffic reports. One in Vermont is testing a driverless bus, as well as sensor networks for truck stop recharging stations.

However, Marr says some of the most interesting innovations are taking place in the cloud, where that will enable “so many more services to be delivered by the car to the home or to the city.”

Automotive-specific clouds are already being used to deliver different services. And IHS Markit predicts the industry will take advantage of cloud computing to bring traditional services like navigation, roadside safety, communication, and diagnostic data online.

One area where AI will take center stage will be connected vehicle services, or “infotainment systems,” as well as newer technologies like mapping and traffic safety. There are already projects underway in Ohio to expand VW’s map-based service by adding AI, road hazards, video, and real-time traffic reports. One in Vermont is testing a driverless bus, as well as sensor networks for truck stop recharging stations.

One technology that will see increasing use in connected vehicle services will be adaptive wireless networks. It enables the car to “run an app, over the network, which then reconfigures the car to be part of a much larger flow of traffic,” Marr says. Sensors, cameras, and other data can be stored on the vehicles’ own servers in order to act as “automated redundancy.”

The effect is not trivial. One Winton Thorne, VP of strategic AI product development at Delphi, which is working on improving adaptive wireless networks, says they allow “distance error reduction” and give the car “environmental feedback.”

Thorne says autonomous driving systems “will operate on a small amount of limited frequency.” But “if the car is in a highly congested lane, the system will know to maintain its speed.” Thanks to the ability to capture traffic-related data from near each vehicle, cars can now take on automatic enforcement. In future, a vehicle can start to manage its own speed.

Currently, automated driving systems can only react to the driver taking his or her hands off the wheel. But when a driver takes his or her hands off the wheel, the system “must retest everything until