Browse
Log in / Register
Share
Date: June 12, 2025 — Every year, thousands of crashes involve vehicles equipped with advanced driver assistance systems (ADAS) such as automatic emergency braking, lane-keeping assist and adaptive cruise control. When one of those crashes happens, the question isn't just who was driving — it's who is legally responsible when the technology itself may have played a role.
The answer, for now, is almost always the driver. But as cars become more capable, insurers, manufacturers and regulators are wrestling with a more complex question: what happens when the line between human error and system failure blurs?
Modern cars can steer, brake and accelerate for extended periods without driver input. Yousif Al-Ani, principal engineer of ADAS at Thatcham Research, a UK-based automotive safety body that tests vehicle systems, says some current systems allow drivers to go more than an hour without touching the wheel or pedals.
"The problem is that they begin to feel like automation," Al-Ani said. "You can have a vehicle that feels autonomous while the driver remains legally liable for everything that happens."
That gap between perception and legal reality is a growing concern. Under the current liability framework, even Level 2 driver-assist systems — which can control both steering and speed simultaneously — leave the driver fully responsible. Regulators in the US and Europe have similarly held that these systems are assistance features, not self-driving technology.
The National Highway Traffic Safety Administration (NHTSA) has opened multiple investigations into crashes involving Level 2 systems, but in most cases the driver is named as the at-fault party. That pattern is unlikely to change unless the vehicle is explicitly certified as an automated driving system.
Even when ADAS functions incorrectly, the driver generally bears responsibility. A false emergency braking event, an unwanted lane-keep correction, or a misunderstanding of system limits all fall under the same rule: the driver must remain attentive and in control.
But insurers don't automatically blame the driver. Thomas Mayo, head of technical underwriting at Allianz UK, said determining liability "requires a case-by-case assessment" that looks at the level of automation in use, whether the driver was expected to intervene, and whether the system performed within its design specifications.
Sian Davies, head of motor liability at Admiral, a UK insurer, said her company's approach is evidence-based. "Our job is to establish what happened based on all available evidence and make fair decisions for our customers," she said.
That evidence increasingly includes data that didn't exist a decade ago. Dashcam footage, telematics records and ADAS event logs can now reconstruct what occurred in the seconds before a crash — sometimes more accurately than witness statements or police reports.
There's a catch: access to that data isn't always available to drivers or insurers.
Al-Ani argues that mandatory vehicle data access is the single biggest change needed to make future insurance claims fair. Without dashcam footage or access to onboard logs, drivers may be presumed at fault by default — even when the system malfunctioned.
"You'll initially be treated as being at fault," he said. That can mean paying a deductible, losing a no-claims discount and facing higher premiums, all because the evidence that would have cleared the driver wasn't accessible.
An Aviva spokesperson said mechanisms for sharing data between manufacturers and insurers "are yet to be agreed." The Association of British Insurers (ABI) has called for a clear data-sharing framework as autonomous vehicle legislation progresses, a position echoed by US groups like the Insurance Institute for Highway Safety and the American Property Casualty Insurance Association.
In the US, the issue is also surfacing in state legislatures. Several states have introduced bills addressing vehicle data access, though federal action has stalled. For now, drivers involved in ADAS-related crashes often face an uphill battle to prove the technology played a role.
The liability landscape will shift when vehicles are certified for approved autonomous operation. Under the UK's Automated Vehicles Act 2024, once an authorized self-driving system is engaged, the driver is no longer legally responsible for the vehicle's actions. Responsibility shifts to the manufacturer or software provider.
Al-Ani says this is a fundamental departure from today's rules. "The driver is no longer legally responsible for the vehicle's actions," he confirmed.
Insurers are already preparing. Mayo said Allianz can "pursue recovery against manufacturers or technology providers where appropriate." Aviva confirms it will still compensate injured parties, even if the automated system is later found to be at fault.
In the US, the picture is more fragmented. There is no federal framework equivalent to the UK's Automated Vehicles Act, and a patchwork of state laws governs testing and deployment. The American Legal Institute has proposed model rules for autonomous vehicle liability, but adoption has been slow. Insurers warn that without a legislative framework and data-sharing standards, liability cases involving self-driving vehicles could be litigated for years.
The core principles of motor insurance are unlikely to change: someone pays, and fault is assigned based on evidence. But as Al-Ani points out, the evidence itself is becoming the battleground. The Automated Vehicles Act provides a legal skeleton, yet questions remain over who can access vehicle data and how quickly it must be produced.
Manufacturers, insurers, regulators and consumer groups are all pushing for clarity — but from different angles. Carmakers generally resist broad data-sharing obligations, citing proprietary technology and cybersecurity concerns. Insurers want access to prove or disprove fault. Consumer advocates want transparency so drivers aren't unfairly penalized.
For now, if you're driving a car with ADAS, you are legally responsible for its operation. The technology may assist you, warn you and even intervene on your behalf, but the law treats you as the person in control.
That will change — but only when cars are genuinely certified to drive themselves and the rules are rewritten around them. Until then, the safest assumption is that the driver pays the price, even when the car makes the call.
