Navigating Autonomous and Electric Vehicle Subrogation

Autonomous vehicles will improve road safety and reduce injuries, death, and property damage. However, casualty and liability trends show that average claim size is increasing due to several factors. In this session at RISKWORLD^® 2026, stakeholders involved in the subrogation of such losses highlighted why these related trends need to be on everyone’s radar.

Speakers included:

• David Thornhill, Senior Vice President of Commercial Claims at CSAA Insurance Group, a AAA Insurer
• Ashton Kirsch, Shareholder at Matthiesen, Wickert & Lehrer, S.C.

Autonomous collisions and associated claims will change with the implementation of autonomous vehicles. The industry is seeing new issues when it comes to autonomous technologies. Product liability typically stems from technology like “autopilot” failing because it was not designed properly. In addition to product failure, there will be a high level of litigation related to decision making and its result. It is expected that verdicts will soar to the billions as a result of these cases.

Auto subrogation will change the following factors forever:

• The nature of auto losses
• The sources of fault
• The parties involved
• The evidence required
• The experts relied upon
• The role of attorneys
• Insurance coverages and recovery

Autonomous technology will not eliminate recoveries. It will reallocate responsibility based on data rather than on testimony. Everything in the field will change as automation is embraced for claimants and defendants alike. It will take years for case law to address all of these factors.

One of the most important changes is how liability will be analyzed. Instead of relying heavily on assumptions or conflicting narratives, the industry is moving toward data-driven conclusions. This shift helps eliminate much of the uncertainty about what actually happened in a loss event. At the same time, loss allocation is becoming more complex. In some cases, responsibility may fall on product liability when an autonomous system fails to perform as designed. In others, human negligence will still play a role, particularly in mixed environments where autonomous and manually driven vehicles share the road.

It is also important to recognize that autonomous technology does not remove driver error entirely. In many cases, it actually makes human error more visible and easier to prove. Fully autonomous systems add another layer of complexity, but they also generate detailed, objective data. This information can either support a product defect claim or clearly demonstrate that the system performed as intended.

From a subrogation perspective, these developments will fundamentally change how fault is established. Investigations will increasingly depend on sensor data, vehicle telemetry, video footage, and advanced event reconstruction. These tools will not just help determine liability, but they also will also be critical in pursuing recovery from autonomous systems when appropriate as well as defending autonomous fleets when they are struck by other parties. Over time, the standard will shift away from competing narratives and toward verifiable performance data.

As product defect claims become more common in this space, claims professionals will need to develop a deeper understanding of when such claims are appropriate. Not every autonomous vehicle loss will involve a product defect, so it is essential to evaluate each case based on the available data. Strong practices in this area include preserving evidence and data as early as possible, conducting timely inspections of the scene and vehicles when warranted, and retaining experts only when the facts truly support their involvement. Maintaining discipline is key. Overreaching product cases can be just as harmful as missing valid recovery opportunities.

Determining responsibility in these cases requires a careful, fact-specific approach. Professionals must consider whether the vehicle was operating autonomously or under manual control, whether human behavior contributed to the incident, and whether the system was functioning within its intended operational design domain. They also need to evaluate whether any component, software, or maintenance failures played a role. Depending on the findings, liability could result from the actions of a human driver, another non-autonomous third party, a manufacturer or supplier, or even a service or maintenance provider. In some cases, the conclusion may be that the autonomous system performed exactly as designed and is not at fault.

Finally, one of the most critical skills in this evolving landscape is the ability to select the right experts. Autonomous vehicle claims are highly specialized. Success often depends on working with professionals who not only understand the technology, but can also interpret complex data accurately. Choosing experts with the right technical knowledge and analytical mindset will be essential as the industry continues to adapt.