The Department for Transport (DfT) has launched a Consultation (the Consultation) on the draft Statement of Safety Principles (SoSP) that will underpin the authorisation and oversight of self-driving vehicles under the Automated Vehicles Act 2024. While the Consultation is primarily aimed at establishing how autonomous vehicles (AVs) will be assessed for deployment on UK roads, it is also highly significant for insurers and risk managers.
The proposed principles provide the clearest indication yet of how government intends to define "safe" automated driving and how the performance of self-driving vehicles will be evaluated over time.
A new benchmark for vehicle safety
At the heart of the consultation is a proposal that self-driving vehicles must perform at a level equivalent to a careful and competent human driver. This is an important policy decision. The standard deliberately exceeds the performance of an average driver, who may be distracted, fatigued or impaired, while remaining capable of operating the vehicle. For insurers, this raises an interesting question: how should liability be assessed where an AV satisfies the regulatory standard but is nevertheless involved in a collision? Future disputes may focus less on whether a vehicle was "perfect" and more on whether it behaved as a competent and careful driver would have done in the same circumstances.
Moving beyond collision statistics
A notable feature of the proposed framework is its emphasis on behaviours rather than outcomes.
The government recognises that, during the early stages of AV deployment, there may be insufficient data to use statistical measures such as collision frequencies or injury rates as the sole determinant of safety performance. Instead, the SoSP focuses on what vehicles should do in real-world situations.
The ten principles require autonomous vehicles to:
Comply with traffic laws and the Highway Code.
Maintain control at all times.
Proactively identify and respond to hazards.
Adapt to road and weather conditions.
Behave predictably.
Improve overall road safety.
Interact safely with vulnerable road users.
Respond appropriately to emergency services.
Operate only within their approved operational design domain.
Account for the specific characteristics of Great Britain's road network.
This behavioural approach mirrors the way liability is often assessed following conventional road traffic collisions, where the actions of a driver are examined against the benchmark of a reasonably careful and competent driver.
Implications for claims and liability
For motor insurers, the Consultation provides valuable insight into the types of evidence likely to become central to future claims investigations. Many of the proposed principles are likely to require extensive vehicle-generated data to demonstrate compliance. Following an accident, questions may include:
Did the system correctly identify a hazard?
Was the vehicle operating within its authorised operational limits?
Did it appropriately recognise roadworks or temporary traffic management?
Could it detect and respond to vulnerable road users?
Did it react appropriately to emergency vehicles?
As a result, claims handlers will inevitably need access to telematics, sensor logs, software records and event data recorder information before they can determine whether an insurers’ liability is triggered under the Automated and Electric Vehicles Act 2018. The importance of data preservation and disclosure protocols is not a matter which has been covered in primary legislation but ought to be addressed in secondary legislation.
Operational Design Domain risks
Principle 9 may ultimately become one of the most significant from a liability perspective. The principle requires vehicles to operate autonomously only within the limits for which they have been designed and authorised. These limits, often referred to as the Operational Design Domain (ODD)—may include restrictions relating to road types, weather conditions, geographic locations, traffic environments and infrastructure capability.
This creates potential new areas of dispute. Claims investigations may increasingly examine whether a vehicle was operating beyond its approved capability, whether warnings were given, and whether responsibility rests with the vehicle, the operator, the software provider or another entity within the AV ecosystem.
Looking ahead
Although the Consultation does not introduce immediate changes to motor insurance obligations, it represents a critical step towards the operationalisation of the Automated Vehicles Act 2024. The Consultation represents an important step forward in the debate on setting the expected standard of the safety of Avs which, in turn, should have a positive impact on accident frequencies and subsequent claims which will be a positive outcome for motor insurers.
The Statement of Safety Principles will ultimately shape how self-driving vehicles are authorised, monitored and assessed throughout their lifecycle.
For insurers, as deployment expands, understanding these principles will be essential for underwriting, claims handling, reserving and the development of new insurance products designed for an increasingly automated transport environment.