How autonomous vehicles will change the future of car insurance

When a Level 3–5 autonomous vehicle is involved in a crash, because system design doesn’t require the driver to be constantly “in the loop”, fault no longer rests with the driver. Instead, liability should shift to the OEM or software provider. This represents a fundamental change in the structure of auto insurance policies.

In response, insurers are developing new insurance products to address varying levels of autonomy and shifting liability. Some policies distinguish coverage based on whether the vehicle was under human control or operating in autonomous mode at the time of an incident. This allows insurers to assign fault based on the responsible party in a given scenario. 

In commercial sectors such as logistics and ride-hailing, some insurers are also exploring coverage frameworks that treat the vehicle as a service-delivery platform. These approaches account for multiple stakeholders, including fleet operators, software vendors, and mobility service providers, who may share responsibility depending on how the vehicle is used.

As automation reduces crashes, fewer claims are expected, but repair costs may rise. AVs contain expensive components such as lidar, sensors, and high-performance processors known as system-on-chip (SoC). The rise of EVs previews this dynamic. While EVs are safer in some respects and mechanically simpler than internal combustion vehicles, they can still be more expensive to repair due to costly parts, limited technician expertise, and tightly integrated systems.

The emergence of robotaxis will further reshape insurance dynamics through changes in fleet composition and vehicle usage patterns. Unlike conventional vehicles, robotaxis are expected to have much shorter lifecycles — around 3–5 years compared to the traditional 12–15 — due to higher annual mileage and faster technology turnover. This creates a more progressive fleet that continually integrates the latest hardware and safety features, potentially reducing accident severity and long-tail risk.

As is already the case today, insurers will rely more on system data than eyewitnesses. Assessing fault may require:

• Existing electronic data recorder (EDR) logs.

• Enhanced EDR with details from camera, radar, lidar, driver monitoring sensors.

• Enhanced EDR with insight on object detection, path planning, and actuation. 

• Timestamped control transitions.

• Software configurations.

• System thresholds and overrides.

As AV systems become more complex, the claims process will require new capabilities to interpret operational data and assign liability accurately. AI is likely to play a leading role in this shift, offering a fast and structured first pass at incident analysis based on sensor logs, vehicle state data, and environmental context.  

Human specialists will still be needed to validate and escalate edge cases, but traditional claims teams may shrink or shift toward higher-level oversight. Photo-estimation has been used for over a decade, and recent advances in AI-driven photo estimating are making the process faster and more reliable. As automation expands, collaboration with OEMs remains essential, yet access to detailed logs and software configurations continues to be a major hurdle.  

One of the biggest hurdles is data access. While insurers could, in theory, require operational data as part of underwriting, this is not yet standard practice — particularly in personal auto contexts. Without regulatory mandates or industry-wide standards, OEMs retain control over vehicle logs and system data — especially true for autonomous vehicles given this form of technology is so recent and constantly evolving. 

As a result, the flow of critical information is often limited or entirely blocked. This makes accurate pricing speculative and claim adjudication significantly more complex, especially when fault attribution hinges on what the vehicle saw, decided, or did.

Deployment is also uneven. AV pilots are active in select cities, but most global regions still rely on traditional vehicles. Roads all over the world will be mixed use, with rare exception, for a very long time. This creates geographic exposure variability that insurers must account for in portfolio planning and product design.