Kodiak AI’s autonomous trucks have successfully driven over 3 million miles, demonstrating the potential for self-driving technology to enhance highway safety in real-world conditions.
Kodiak AI, a prominent player in the field of AI-powered autonomous driving technology, has been quietly proving the viability of self-driving trucks on actual highways. The company’s flagship system, known as the Kodiak Driver, integrates advanced software with modular, vehicle-agnostic hardware, creating a cohesive platform designed for the complexities of real-world trucking.
As Kodiak AI explains, the Kodiak Driver is not just a theoretical solution; it is built to address the challenges of highways, varying weather conditions, driver fatigue, and the demands of long-haul transportation. This practical approach is essential, as trucking is far from a controlled laboratory environment.
In a recent episode of CyberGuy’s “Beyond Connected” podcast, Kurt spoke with Daniel Goff, vice president of external affairs at Kodiak AI, about the evolving perceptions surrounding autonomous trucks. Goff reflected on the initial skepticism the company faced when it was founded in 2018. “When I first started at the company, I said I worked for a company that was working to build trucks that drive themselves, and people kind of looked at me like I was crazy,” he recalled. However, he noted a significant shift in public sentiment as autonomous vehicles have begun to demonstrate their capabilities beyond mere hype.
One of Kodiak AI’s key arguments is that machines can mitigate many risks associated with human driving. Goff emphasized, “This technology doesn’t get distracted. It doesn’t check its phone. It doesn’t have a bad day to take it out on the road. It doesn’t speed.” In the trucking industry, where safety is paramount, these “boring” characteristics of autonomous vehicles can be advantageous.
Kodiak AI has been actively operating freight routes for several years, rather than solely conducting tests in controlled environments. The company has a command center in Lancaster, Texas, which has facilitated deliveries to cities such as Houston, Oklahoma City, and Atlanta since 2019. During these operations, a safety driver is present to take control if necessary, allowing Kodiak to refine its technology in real-world conditions.
Long-haul trucking is crucial to the U.S. economy, yet it is also one of the most demanding and hazardous professions. Drivers often spend extended periods away from home, working long hours while managing heavy vehicles under various conditions. Goff pointed out that the job’s challenges are compounded by federal regulations that limit driving hours to reduce fatigue. “Driving a truck is one of the most difficult and dangerous jobs that people do in the United States every day,” he said. With a growing number of drivers retiring and fewer individuals entering the profession, the industry is experiencing a significant driver shortage.
Kodiak AI believes that autonomous technology is best suited for the most challenging and repetitive tasks within trucking. Goff explained, “The goal for this technology is really best suited for those really tough jobs—the long lonely highway miles, the trucking in remote locations where people either don’t want to live or can’t easily live.” He also noted that many trucks are idle for a significant portion of the day, with the average truck being driven only about seven hours daily. Autonomous technology could help optimize this by enabling trucks to operate around the clock, only stopping for refueling and safety inspections.
With over 3 million miles driven, Kodiak AI has established a strong safety record, with a safety driver present for most of those miles. Goff highlighted the scale of their operations by comparing it to the average American’s lifetime driving distance of approximately 800,000 miles. “We’re at almost four average lifetimes with our system today,” he stated. The company also utilizes computer simulations and various assessments to evaluate the safety of its system.
In addition to long-haul operations, Kodiak AI collaborates with Atlas Energy Solutions for oil logistics in the Permian Basin. As of the third quarter of 2025, the company has delivered ten driverless trucks to Atlas, which autonomously transport sand around the clock without a human operator in the cab. Goff described this partnership as an ideal environment for testing and refining their long-haul operations.
Kodiak AI has sought third-party validation of its safety claims, including a study with Nauto, a leader in AI-enabled dashcams. The results indicated that Kodiak’s system achieved the highest safety score recorded by Nauto.
Policy and regulation also play a critical role in the adoption of autonomous trucking. Goff noted that 25 states have enacted laws allowing for the deployment of autonomous vehicles. He believes that the inherent dangers of driving make a compelling case for the technology. “People who think about transportation every day understand how dangerous driving a car is, driving a truck is, and just being on the road see the potential for this technology,” he said.
Despite the advancements, concerns about safety remain prevalent among advocates and everyday drivers. Critics question whether autonomous systems can respond adequately in emergencies or handle unpredictable human behavior on the road. Goff acknowledged these concerns, stating, “In this industry in particular, we really understand how important it is to be safe.” He emphasized that trust in autonomous systems must be earned through consistent real-world performance and transparent testing.
For everyday drivers, the prospect of sharing the road with autonomous vehicles can be unsettling, especially given the focus on potential failures in media coverage. However, Kodiak AI argues that the removal of human factors such as fatigue and distraction could lead to safer highways. If the technology continues to perform as claimed, it could result in fewer tired drivers on overnight routes, more reliable freight movement, and ultimately safer roads for all users.
As Kodiak AI continues to move freight and gather safety data on public roads, skepticism remains a vital aspect of the conversation surrounding autonomous trucking. The future of this technology will depend on its ability to demonstrate long-term safety benefits and earn the trust of the public, rather than relying on promises alone. The pressing question is no longer whether self-driving trucks can operate effectively, but whether they can consistently prove to enhance safety for everyone on the road.
For further insights, refer to CyberGuy.