Personal Injury Claims Involving AI-Powered Vehicles: Who Is Responsible?

Artificial intelligence is transforming the way we move around, and autonomous and semi-autonomous technology is becoming commonplace in more vehicles on Australian roads, with features such as adaptive cruise control, lane-keeping assistance and automated emergency braking no longer limited to premium models and rolling out to a wider range of new cars. While this technology is designed to prevent collisions and enhance road safety, accidents involving AI-powered vehicles raise complex legal questions. While a traditional motor vehicle accident may involve assigning fault based on driver conduct, in many cases, a crash involving automated technology may necessitate a careful examination of the actions of the driver. The examination may also involve the roles of vehicle manufacturers, software developers, and component suppliers. This technology is still developing, and personal injury claims are becoming increasingly complex, both technically and legally.

The Growing Presence of Automated Driving Technology

A number of state and federal initiatives have promoted testing and development of automated vehicle technology. While fully autonomous vehicles are still rare, partially automated driving systems are beginning to appear on public roads more and more frequently. The worry is that many drivers do not understand what these technologies can and cannot do. In 2024, IIHS released research showing minimal evidence that partial automation systems lower crash rates beyond what is already being achieved by current technologies such as automatic emergency braking.

That result contradicts the popular misconception that AI-enhanced driving makes roads safer and highlights the dangers of crashes that occur when drivers rely too heavily on automated systems. In the context of personal injury litigation, that distinction has significant legal significance, as courts may consider whether it was reasonable for a driver to rely on an automated system in the circumstances. Matters involving determining liability are increasingly being handled by legal professionals, such as personal injury lawyers Sydney CBD. Liability may be determined by analysing both human decision-making and the behaviour of vehicle software.

When the Human Driver Remains Responsible?

While vehicle automation continues to advance, Australian road laws typically still place legal responsibility on the driver. At SAE Level 2 automation, the vehicle can assist with steering, acceleration and braking, but it still requires the driver to continuously monitor the road and be prepared to assume control.

In fact, a multi-year review by the United States National Highway Traffic Safety Administration (NHTSA) found that at least 13 fatal crashes occurred while Tesla’s Autopilot system was in use, with regulators concluding that in some cases, inadequate driver engagement and overreliance on automation were to blame. These findings are indicative of why courts will likely still hold drivers responsible for collisions when they are not paying attention to the road, even if an automated driving system is engaged. When it comes to personal injury proceedings, all of this evidence, such as vehicle log data, driver monitoring records, and information captured by event data recorders, may help determine whether the driver acted reasonably up until the collision occurred.

Manufacturer Liability and Product Defects

Vehicle manufacturers could also face liability if a crash can be linked to a defect in the automated driving system, such as sensors that fail, software that does not correctly identify hazards, or system warnings that are not strong enough to alert the driver.

These risks play out in practice, as we can see from recent regulatory investigations. In 2024, federal regulators investigated 22 separate incidents in which Waymo self-driving vehicles collided with stationary objects and traffic-control devices. Although many of these incidents did little more than cause minor damage, they led to questions about whether autonomous systems can properly read complex road environments. Under established product liability principles, manufacturers can be held liable for large amounts of compensation if a personal injury claim establishes that a design defect directly contributed to a crash.

The Challenge of Software and Algorithm Accountability

Unlike traditional product defects that involve mechanical failures, accidents in AI-powered vehicles may be due to a decision made by software rather than a mechanical component. It is much more difficult to determine legal responsibility when algorithms are constantly analyzing environmental information and making driving decisions in real time.

In a 2026 analysis of over 2,168 autonomous vehicle incidents reported to NHTSA, perception and planning failures were the main cause in about 64% of the incidents investigated. About 50% of the incidents involved rear-end collisions. Software-related failures are a major factor in autonomous vehicle crashes, and expert testimony may be needed to determine whether the software functioned as it was intended to or if defects in its programming caused the accident.

Shared Liability Between Multiple Parties

Some accidents involving AI-powered vehicles will likely not lead to one party being solely liable; a driver may fail to act when necessary, while the automated driving system makes a mistake. The investigation may involve manufacturers, software developers, sensor suppliers, and fleet operators.

This approach to shared liability differs markedly from a conventional motor vehicle claim. Courts may ultimately apportion responsibility between multiple parties after considering detailed technical evidence. Resolving these disputes frequently requires comprehensive accident reconstruction, in-depth software analysis, and expert testimony. As automated vehicle technology continues to develop and the surrounding ecosystem becomes increasingly sophisticated, multi-party personal injury litigation is expected to become progressively more common across Australia and other developed jurisdictions.