Automated vehicles: statement of safety principles consultation

Question 1

Introduction

Accepted Answer

The Automated Vehicles Act 2024 marks a pivotal step in shaping the regulatory framework for self-driving vehicle technology in GB. This legislation:

establishes the legal foundation for the safe deployment and use of self-driving vehicles also known as automated vehicles (AVs) on public roads
supports innovation while prioritising public safety, transparency and accountability
builds on a 4 year AV review by the Law Commission of England and Wales and the Scottish Law Commission (the Law Commissions)

The act introduces a comprehensive safety framework, which creates a whole life ‘in-use’ assurance system for AVs consisting of:

type approval which evaluates whether the vehicle meets all technical safety, environmental and other performance standards
an authorisation process that will ensure that the vehicle can drive safely and legally on our roads without a driver
ongoing regulation with powers to monitor safety data and enforce compliance, as well as an independent investigation function

At the heart of this framework is the statement of safety principles (SoSP). This will set out the safety standard to which AVs will be expected to comply.

Call for evidence

We ran a call for evidence between 10 June 2025 to 1 September 2025 on the SoSP. We sought information on how:

the safety principles may be used
the safety standard may be described
safety performance could be measured

Responses to the call for evidence have informed the development of the draft statement of safety principles.

Question 2

The purpose of the SoSP

Accepted Answer

Under the AV Act 2024 the SoSP has 2 statutory purposes:

under section 1 the Secretary of State for Transport must have particular regard to the SoSP when considering at authorisation whether vehicles can travel safely and autonomously
under section 38 the Secretary of State must also assess the annual performance of the self-driving fleet, including the extent to which that performance is consistent with the SoSP

The primary focus of the SoSP is to fulfil its statutory purpose of providing guidance to the Secretary of State in considering the self-driving test at authorisation.

Section 1 of the AV Act sets out that a vehicle ‘satisfies the self-driving test’ if:

it is designed or adapted with the intention that a feature of the vehicle will allow it to travel autonomously
it is capable of doing so, by means of that feature, safely and legally

The principles have therefore been drafted to focus on the vehicle and its features in the context of meeting the self-driving test.

Details of the authorisation process will be subject to consultation in due course.

This does not prevent the principles having wider uses within the AV safety framework. Such wider uses might include the in-use regulator taking the SoSP into account when deciding whether to investigate an incident or issue a sanction.

The focus of the SoSP on whether a vehicle meets the self-driving test means that it does not address safety through the transition of handing control of the vehicle back to a user-in-charge, which will be addressed through the wider AV regulatory framework.

The call for evidence sought views on the potential uses of the SoSP and elicited a range of alternative purposes in addition to these statutory ones. Suggestions included:

providing a tool for public education and communication to provide assurance on safety
supporting insurance framework by providing clear definitions of acceptable risk levels for AVs
extending the scope of the safety principles to cover the supply chain, third party providers or ongoing safety such as requiring software updates

Since the call for evidence was published, the Department for Transport (DfT) has published a call for evidence on Developing the automated vehicles regulatory framework. This sets out some emerging thinking regarding how the wider safety framework might operate and provides some of the detail that respondents to the SoSP call for evidence sought on issues such as cyber security or the operation of the in-use regulator.

We consider many of the issues identified to be better addressed through other aspects of the comprehensive safety framework, rather than the SoSP.

Question: do you agree or disagree with the purpose of the SoSP being to provide guidance to the Secretary of State in considering the self-driving test?

Question: why do you think this, and, if relevant, what alternatives would you prefer?

Question 3

Setting the safety standard

Accepted Answer

Road safety and AVs

In January 2026, the government published a road safety strategy which sets out a clear and ambitious path to improve road safety in Great Britain. It introduces targets to reduce by 65% the number of people killed or seriously injured on roads in Great Britain by 2035, using a 2022 to 2024 baseline.

At the heart of this road safety strategy is the safe system. The safe system is an internationally recognised evidence-based approach to road safety. It recognises that while human error is inevitable, deaths and serious injuries are not; the road and the vehicle environment should be designed to protect people as much as possible.

One of the themes within the road safety strategy is taking advantage of technology, including recognising the potential of safe AVs to have road safety benefits.

According to Self-driving vehicles: new safety ambition, in 2020 88% of all recorded collisions on roads in GB involved human error as a contributory factor. AVs have the potential to reduce the number of people killed or seriously injured through their training on extensive driving data combined with sensor capabilities that may enable them to identify potential hazards earlier and react faster to a hazard than human. Unlike human drivers, AVs don’t get tired, distracted or drive under the influence.

However, AVs are different from humans. We should expect some instances where the capabilities exhibited by an automated vehicle when executing elements of the driving task may look and feel different to a human. This may introduce different risks, especially when operating in a mixed traffic environment. It would therefore be unrealistic to expect that AVs will have zero collisions.

Safety is paramount to the AV regulatory framework, of which the SoSP is one part.

Question 4

The UNECE automated driving systems (ADS) regulation

Accepted Answer

In addition to the domestic regulatory framework, the United Nations Economic Commission for Europe (UNECE) automated driving systems (ADS) regulation has been developed internationally which:

establishes uniform safety provisions and a harmonised methodology for validating automated driving system safety to enable the safe introduction of this technology in the market.

AVs will be scrutinised against the requirements within the regulation as part of the type approval process. This includes not only vehicle features but also a much broader focus on manufacturer safety processes and culture, which combines ADS performance requirements and safe user interactions with organisation‑level obligations for:

a safety management system, including in-service monitoring and reporting on the performance of the ADS
an evaluation of the test methods used (simulation/track/real-world testing)
a structured safety case with claims, arguments and evidence

All these measures are aimed at demonstrating the ADS is free from unreasonable risk.

If adopted, the ADS regulation is expected to come into force in early 2027.

The ADS regulation includes ‘careful and competent’ as a safety standard, stating that:

As a general concept, this Regulation requires the ADS to deliver a level of safety at least equivalent to that of a competent and careful human driver.

We favour alignment with international standards such as the ADS regulation. International harmonisation establishes a common safety baseline and shared assessment methods, which:

lowers market‑entry costs
enables cross‑border deployments
facilitates faster spread of safety technologies

Considering the safety standard

The AV Act requires that the SoSP be framed with a view to securing that:

authorised automated vehicles will achieve a level of safety equivalent to or higher than, that of careful and competent human drivers
road safety in Great Britain will be better as a result of the use of authorised automated vehicles on roads than it would otherwise be

There has been significant consideration of the safety standard in recent years. The Law Commission and Scottish Law Commission published a Automated vehicles: joint report in January 2022 outlining recommendations for legal reform around AVs and how they could safely and responsibly be introduced on GB roads and public places.

Their report recommended that the UK government should publish a safety standard against which the safety of AVs can be measured and which would inform the regulatory regimes of authorisation, in-use regulation and no user-in-charge licensing. The safety standard would help to reassure the public that AVs are safer than human drivers and provide a basis on which to take regulatory action where an AV’s performance falls below an acceptable threshold.

In thinking about the safety standard, the Law Commission proposed 3 standards:

As safe as a competent and careful human driver
As safe as a human driver who does not cause a fault accident
Overall, safer than the average human driver

When these options were consulted on, the Law Commissions found that all respondents agreed that AVs should be safer than human drivers in general - but thereafter opinions divided. Given their divergent views, the Law Commission recommended that legislation should require the Secretary of State for Transport to publish a safety standard against which the safety of AVs can be measured.

In Connected and automated mobility 2025: realising the benefits of self-driving vehicles the last government set out its plans for connected and automated mobility technologies and stated within it that safety is the first priority in developing AVs. It went on to propose a safety standard:

This government believes that self-driving vehicles should be held to the equivalent standard of behaviour as that expected of human drivers; competent and careful.

In response to this consultation, 56% respondents were in favour of this as a safety standard and 39% were against. Several respondents sought a higher level of ambition and others expressed concern that AVs would not be able to meet a careful and competent standard or would not result in a reduced number of collisions. Many respondents sought clarity on the definition of a careful and competent human driver and how it applies to an AV.

The government ran a call for evidence on the safety principles between June and September 2025. This sought views on the characteristics and implications of a safety standard set at either careful or competent or higher than careful and competent.

The call for evidence identified that there were a range of views regarding the appropriate level at which the safety standard might be set. The full summary of responses has been published.

Key points made by respondents about the safety standard included:

When asked about what characterises careful and competent human driving, respondents identified a range of behaviours, such as good vehicle control, hazard perceptions, obeying traffic laws, interacting well with other road users and adapting to driving conditions.
When asked about the characteristics of a higher than careful and competent human driver, respondents identified behaviours, such as earlier detection of hazards, adaptability to unexpected circumstances and better safety outcomes.
A number of respondents felt that equivalent to careful and competent human driver was too low as a standard. It was not seen as a high standard for human drivers, represented an unambitious standard for AVs given their capabilities and may reduce public acceptance of the technology. Others felt that careful and competent represented a higher standard than an average human driver and represented an appropriate balance between achieving an appropriate level of safety and realising the early deployment of AVs.
Some respondents felt that a standard higher than careful and competent would result in road safety improvements. However, a number also felt that it would have a detrimental impact on developers, which may delay deployment and mean that the benefits of AVs are reduced or felt more slowly.
One respondent noted that a higher standard may result in more aggressive driving from human drivers if an AV was felt to drive too cautiously. Industry respondents also noted that a standard higher than careful and competent would not align with emerging international regulations and was not realistic or measurable.

In their report Safety first? Understanding acceptance of automated vehicles DG Cities found that, while safety was not a priority for many respondents to the survey, elevating AV safety to a level above careful and competent human drivers had the potential to significantly increase public support for AVs. They also noted in the report that direct messaging that emphasized AV safety features was more effective than indirect messaging that highlighted human driver error.

DfT commissioned deliberative research into the public’s perception of the safety standard. The research sought to explore the public’s understanding of careful and competent and whether it represented an appropriate safety standard, through 6 workshops across 3 locations within the UK.

Key points from this research include:

There was moderate positive support for careful and competent as a benchmark. People were more accepting of careful and competent as a benchmark when it was clear that it was a higher standard than that of an average driver.
Participants often felt that ‘careful and competent’ was somewhat ambiguous and may not be easily understood without further explanation.
Participants tended to describe careful and competent driving in behavioural terms rather than statistical ones or with reference to outcomes. One suggestion was to use the advanced driving exam as a good benchmark which was positively regarded by other participants.

Question 5

Setting the safety standard in the SoSP

Accepted Answer

A careful and competent human driver represents a higher standard than an average driver. ‘Average drivers’ includes those who may have the necessary skills to be competent but may not always be careful and those who are impaired, distracted or intoxicated. Deploying AVs designed to this standard is expected to improve safety outcomes.

Peer reviewed research conducted by Waymo in the USA shows that until January 2025, taking into account 56.7 m rider-only miles without a human driver behind the wheel, their fleet caused an 85% reduction in serious injury crashes.

They also claim 92% fewer pedestrian crashes with injuries, 82% fewer cyclist crashes with injuries and 82% fewer motorcycle crashes with injuries than a human driver.

This evidence suggests that current AV deployments, although not in the UK, are, on average, performing better than an ‘average human driver’. There is also evidence that one operator is comparing their model to an ‘ideal human driver’, which does not get distracted or fatigued, which may resemble a ‘careful and competent’ driver, under some definitions. However, these findings are specific to one AV operator and may differ across the industry and any future deployments in the UK.

Research on insurance claims relating to Waymo vehicles, based on peer reviewed methodology and using data provided by Waymo supports this further by finding that over 25.3 million miles with the Waymo ADS “significantly outperformed both the overall driving population” with reductions of 88% in property damage claims and 92% in bodily injury claims.

There is no definition of a safety standard set at higher than a careful and competent human driver. In particular, industry respondents to the call for evidence highlighted that there was no defined or measurable standard for what a higher than careful and competent standard would look like.

While respondents to the call for evidence identified a number of behaviours they considered to be consistent with higher than careful and competent, many of these were consistent with the behaviours suggested for careful and competent.

As the draft ADS Regulation recognises careful and competent as a safety standard, there is also a risk that a standard higher than that will result in requirements being placed on developers in addition to those specified internationally.

This is likely to impact on the speed and efficiency with which authorisation can be undertaken. Consequently, we would anticipate this to delay the deployment of AVs on GB roads and with that the delivery of expected road safety benefits.

We have proposed within the draft SoSP that the safety standard be set at equivalent to careful and competent human drivers. As we consider it represents an appropriate balance between ensuring road safety and not preventing the realisation of safety benefits through the deployment of AVs.

Question: Do you agree or disagree with the safety standard for automated vehicles being set at equivalent to careful and competent human drivers?

Question: why do you think this, and, if relevant, what alternatives would you prefer?

Question 6

Using behaviours within the SoSP

Accepted Answer

The call for evidence did not offer a fixed definition of what careful and competent is, although suggested some broad characteristics that might be associated with each. When asked about careful and competent driving, most respondents to the call for evidence identified behaviours that they associated with it.

Participants in the public perception of the safety standard workshops similarly “were readily able to describe careful and competent in terms of behaviours rather than expected driving outcomes”.

Other bodies are also considering careful and competent driving in relation to a series of behaviours. The Driving Vehicle and Standards Agency (DVSA) have responsibility for carrying out driving tests and approving people to be driving instructors.

Working closely with the Vehicle Certification Agency (VCA), who are a designated UK Type Approval Authority for automotive products, they have developed a set of behavioural competencies for AVs. This utilises the assessment framework for driving examiners and applied it to AV behaviour.

The framework is intended in the first instance as an assessment framework for considering applications for a vehicle special order in support of commercial piloting, although this is expected to support assessment at authorisation when the wider regulatory framework is implemented.

An approach of considering ‘careful and competent’ as behaviours is emerging in other countries. The Netherlands vehicle authority, RDW, is the Dutch type approval authority. They are considering an assessment approach to automated lane keeping systems (ALKS).

They define a ‘careful and competent driver’ as a having a level of driving proficiency equivalent to that of an existing vehicle driver who qualifies for admission to the CBR’s examiner training programme. The assessment compares the driving behaviour against various categories which draw on various human driving standards.

We consider that if a vehicle demonstrates behaviours consistent with careful and competent human drivers, then AVs collectively will be equivalent to careful and competent and that this will result in an overall improvement in road safety.

The SoSP has therefore focused on identifying behaviours equivalent to careful and competent that can be tested as part of the self-driving test at authorisation.

While the focus of the SoSP may be on pre-deployment, the focus on behaviours would also provide the Secretary of State with a useful framework for assessing whether a collision or driving behaviour after deployment meets the appropriate standard and could inform any decision relating to sanctions.

Question: Do you agree or disagree with using behaviours as the basis for developing the principles within the statement of safety principles?

Question: why do you think this, and, if relevant, what alternatives would you prefer?

Question 7

Identifying careful and competent behaviours

Accepted Answer

No clear definition of careful and competent driving emerged through the call for evidence responses. Respondents to the call for evidence typically identified behaviours that they associated with careful and competent, such as:

good vehicle control, predictable driving, smooth driving which avoids harsh braking and cornering and not driving impaired or under the influence
hazard perception, such as successfully perceiving hazards and reacting to them, being able to anticipate hazards and take action to avoid them, being aware of your surroundings and making context-specific evaluations of risks and hazards
obeying all road traffic laws and safety-critical road signs, including the Highway Code
successfully interacting with other road users – this included being aware of other road users and their actions, being able to communicate effectively with other road users, including understanding hand signals, being able to anticipate the actions of other road users and paying particular attention and care to vulnerable road users
the ability to account for and drive safely in, all environmental conditions as a key characteristic

Similarly, participants in the public perceptions of safety research tended to identify careful and competent behaviours such as anticipating hazards, maintaining awareness of surroundings, driving at appropriate speeds for conditions and interacting safely with other road users. A full list of the behaviours identified is set out in their research report.

Across these different sources it is possible to identify 5 broad themes:

vehicle control
hazard anticipation, perception and reaction
obeying traffic rules
interaction with other users
adaptability to road conditions

From these we have developed a principle (or principles) to cover each area.

However, respondents to the SoSP call for evidence also identified several behaviours associated with AVs only, such as:

continuous learning and improvement at a system level through sharing learning across the AV fleet
ability to detect and respond to faults within the vehicle
transparent and risk-based decision taking

We recognise that there are some AV specific attributes which need to be reflected in the SoSP and have developed principles relating to 2 of these:

compliance with the limits of its authorisation, in terms of geography and capability of the vehicle
designing vehicles to consider the specificities of the territory in which they operate

Question: do you agree or disagree with the identified behaviours of:

vehicle control
hazard anticipation, perception and reaction
obeying traffic rules
interaction with other road users
adaptability to road conditions

Question: why do you think this?

Question: in your view, if any, what behaviours do you think are missing and why?

Question 8

Safety outcomes

Accepted Answer

The AV Act 2024 identifies 2 safety outcomes for AVs:

authorised automated vehicles will achieve a level of safety equivalent to or higher than, that of careful and competent human drivers
road safety in Great Britain will be better as a result of the use of authorised automated vehicles on roads than it would otherwise be

These outcomes will be measured and assessed as part of the general monitoring duty on the Secretary of State to monitor the performance of the self-driving fleet, including against the SoSP.

One suggestion, from both a respondent to the call for evidence and a participant in the public perceptions research workshops was to set a numerical safety standard. For example, the standard could be set so vehicles have:

x% fewer collisions than human drivers
perform at a level similar to the safest y% of human drivers

In considering whether the SoSP could include a numerical target, we have taken into account a range of factors.

A numerical target would require large volumes of evidence to be able to demonstrate the different safety outcomes. Authorisation tests whether vehicles can drive safely and legally. A wide range of evidence sources can be gathered for this, including within simulation, track-testing and real-world testing.

AV deployments are currently at an early stage and there are currently no deployments at scale of self-driving vehicles in GB. Evidence can be gathered by AV developers of safe performance on GB roads through use of the trials under the Code of Practice or through driverless pilots.

Due to their small scale and limited geographical scope, we do not anticipate the scale of deployments will enable the collection of high enough volumes of data to be able to draw statistically significant conclusions on whether a numerical safety standard had been achieved in the short-medium term.

Another consideration is how to set an appropriate benchmark against which to measure the safety outcomes delivered by AVs. Any benchmark would need to strike the right balance between achieving near-term benefits and setting a bar too high that it cannot be achieved.

Challenges in setting an appropriate benchmark include the varying safety levels across different types of roads and collection of sufficient data for localities. Departmental statistics may provide useful data relating to collisions but may not represent the full range of collisions in Great Britain.

In previous years there have also been unexpected issues with data collection for some police forces. This has meant that the reliability and accuracy of the recorded road collisions in some regions cannot be guaranteed.

Whilst this is unlikely to noticeably impact on the overall Great Britain road collision trends, it will impact reliability of comparisons at a local level.

In any event, measurement of road collisions or KSI statistics is necessarily a lagging metric; whether the metric has been met or not will not be known to the Secretary of State at the time of authorisation in the short to medium term.

The draft SoSP has not set any additional safety outcomes beyond those specified in the act.

Question 9

Overall improvement in road safety

Accepted Answer

We expect automated vehicles will achieve an overall improvement in road safety if a level of safety equivalent to careful and competent human drivers is developed and deployed. However, we believe the overall improvement in road safety should not come at the expense to safety of any groups of road users.

In the call for evidence, we proposed to include an equalities and fairness safety principle that focuses on the outcomes between different groups of road users, such as cyclists and pedestrians. Respondents to the call for evidence were largely supportive of this approach.

To achieve this outcome, a principle has been included within the draft SoSP that aims to ensure that road safety is improved for all and that risks are minimised to all road users. In the event that risks do occur, the principle seeks to ensure that potential for harm is minimised.

Additionally, there is a principle that seeks to ensure that AVs interact safely with all road users, particularly those most at risk of collision.

Question: do you agree or disagree that the proposed principles on improving road safety for all and interacting safely with other road users provide sufficient protection for all road users?

Question: why do you think this, and, if relevant, what alternatives would you prefer?

Question 10

The draft safety principles

Accepted Answer

The draft SoSP proposes 10 principles which are outlined below.

1. Self-driving vehicles should use the road in accordance with traffic laws and the Highway Code

Self-driving vehicles should comply with road traffic laws, speed limits, signage, road markings and any temporary restrictions such as roadworks or diversions.

2. Self-driving vehicles should maintain control over the vehicle at all times.

This includes performing the driving task smoothly, safely and in a stable manner, including during unexpected events or rapidly changing conditions.

3. Self-driving vehicles should predict, detect and respond to hazards proactively

This includes predicting and detect potential or existing hazards and take safe, timely and contextually appropriate action to avoid or minimise harm.

4. Self-driving vehicles should adapt driving to road and weather conditions.

This includes recognising the conditions of the road and weather and adjust its driving strategy accordingly, to reflect the environment and maintain safe margins.

5. Self-driving vehicles should behave in a predictable manner

This includes behaving in a way other road users understand and communicating its intention through appropriate signals to other road users clearly.

6. Self-driving vehicles should improve road safety in Great Britain for all

This includes seeking to minimise risk to its occupants and other road users, avoiding any disproportionate increase in risk to any group of road users. When safety risks occur the self-driving vehicle should seek to minimise the potential for harm.

7. Self-driving vehicles should interact safely with all road users, particularly vulnerable road users.

This includes, though is not limited to, being considerate around pedestrians, cyclists, motorcyclists, children, older adults and disabled people more at risk.

8. Self-driving vehicles should interact safely with emergency services

This includes detecting and responding safely to instructions from emergency services and other authorised officials.

9. Self-driving vehicles should only operate autonomously within the limits under which they are specifically designed to function.

This means recognising its limits (including the capability of the automated driving system to perform the driving task) and having appropriate strategies in place to respond where there is a risk that a vehicle may exceed those limits.

10. Self-driving vehicles should be designed to take account of the specificities of the territory where they will operate

This includes when considering whether a vehicle travels safely by means of comparisons to human drivers or using data derived from human drivers, such data and comparisons should be:

representative of comparable locations and circumstances
sufficiently representative of GB drivers who are unimpaired, not distracted and follow traffic rules

Question: Do you agree or disagree with the inclusion of Principle 1: self-driving vehicles should use the road in accordance with traffic laws and the Highway Code?

Question: Do you agree or disagree with the inclusion of Principle 2: self-driving vehicles should maintain control over the vehicle at all times?

Question: Do you agree or disagree with the inclusion of Principle 3: self-driving vehicles should predict, detect and respond to hazards proactively?

Question: Do you agree or disagree with the inclusion of Principle 4: self-driving vehicles should adapt driving to road and weather conditions?

Question: Do you agree or disagree with the inclusion of Principle 5: self-driving vehicles should behave in a predictable manner?

Question: Do you agree or disagree with the inclusion of Principle 6: self-driving vehicles should improve road safety in Great Britain for all?

Question: Do you agree or disagree with the inclusion of Principle 7: self-driving vehicles should interact safely with all road users, particularly those most at risk?

Question: Do you agree or disagree with the inclusion of Principle 8: self-driving vehicles should interact safely with emergency services?

Question: Do you agree or disagree with the inclusion of Principle 9: self-driving vehicles should only operate autonomously within the limits under which they are specifically designed to function?

Question: Do you agree or disagree with the inclusion of Principle 10: self-driving vehicles should be designed to take account of the specificities of the territory where they will operate?

Question: Do you have any comments, including textual amendments, on Principle 1?

Question: Do you have any comments, including textual amendments, on Principle 2?

Question: Do you have any comments, including textual amendments, on Principle 3?

Question: Do you have any comments, including textual amendments, on Principle 4?

Question: Do you have any comments, including textual amendments, on Principle 5?

Question: Do you have any comments, including textual amendments, on Principle 6?

Question: Do you have any comments, including textual amendments, on Principle 7?

Question: Do you have any comments, including textual amendments, on Principle 8?

Question: Do you have any comments, including textual amendments, on Principle 9?

Question: Do you have any comments, including textual amendments, on Principle 10?

Question 11

Principles that we haven’t included

Accepted Answer

The call for evidence identified 3 principles which we considered did not need to be an explicit principle within the SoSP. These were:

the ability to drive without human monitoring of the vehicle or road environment or without human control.
cyber resilience
explainability

While the ability to drive without human monitoring is not expressly stated in the SoSP, it can reasonably be understood as implicit within it. Taking into account the responses received to the call for evidence, we continue to believe that cyber resilience and explainability are best dealt with elsewhere. Our emerging approach to cyber security is set out in Developing the automated vehicles regulatory framework.

We also considered a range of other behaviours for inclusion in the safety principles. These included:

managing risks to passenger safety, such as ensuring that seatbelts are used
passenger wellbeing
ensuring that software updates are undertaken
participating fully in incident investigations
learning from incidents is shared and improvements made

These have not been taken forward into the SoSP as other parts of the regulatory regime are the appropriate place to manage these safety risks. For example, authorisation conditions can ensure authorised self-driving entities keep vehicles up to date and share learning.

NUICO licensing will have general responsibility for problems arising during a NUIC journey overseen by the operator. The AV act provides a range of powers to monitor safety data and enforce compliance through an in-use regulatory scheme, as well as an independent investigation function.

Question: in your view, should any currently excluded areas be included in the statement of safety principles?

Question: which areas should be included? Why do you think they should be included?

Question 12

Data and monitoring

Accepted Answer

In the call for evidence, we also sought views on metrics for monitoring the performance of the AV fleet under the general monitoring duty in the AV Act 2024.

There were many responses concerning the metrics that might be adopted for monitoring, including leading and lagging metrics and the evidence sources that might be utilised to report on these metrics. There were also many comments regarding the process of collecting, auditing and publishing data gathered.

The Secretary of State for Transport must report on the performance of the AV fleet annually and consideration will be given to the comments received as part of the call for evidence in developing the monitoring framework.

The draft ADS Regulation proposes a system of data gathering in relation to incidents with AVs. The AV Act 2024 also contains a range of powers relating to the ability to request or require that data is provided and civil or regulatory sanctions to address any regulatory failures. We intend to continue to develop our thinking on the use of these powers and approach to monitoring as we develop the regulatory framework.

Question 13

Costs and benefits

Accepted Answer

Some in society are likely to benefit from new legislation, while others may see increased costs. These costs and benefits could affect businesses, households, government and wider society. We are interested in understanding what the impact of the SoSP will be on you, your organisation or wider society. Responses to the following question will be considered when conducting appraisal of policy options for the SoSP.

Question: In your view, what costs should be taken into consideration when assessing the impact of the statement of safety principles?

Question: In your view, what benefits should be taken into consideration when assessing the impact of the statement of safety principles?

Question 14

The draft statement of safety principles

Accepted Answer

Purpose

Under the Automated Vehicles (AV) Act 2024 the Secretary of State must prepare a statement of safety principles for the purpose of assessing whether a vehicle is capable of travelling autonomously and safely. This assessment forms part of the authorisation process before such vehicles can legally be deployed on roads or other public places in Great Britain without being monitored by a human driver.

The statement has been prepared to fulfil this statutory function.

Setting the safety standard

Under the AV act 2024 the safety principles must be framed with a view to securing that:

authorised self-driving vehicles will achieve a level of safety equivalent to, or higher than, that of careful and competent human drivers
road safety in Great Britain will be better as a result of the use of authorised self-driving vehicles on roads than it would otherwise be

The Secretary of State has set the safety standard at equivalent to that of careful and competent human drivers.

This is the minimum standard which self-driving vehicles are expected to meet.

Monitoring

The Secretary of State must also put in place effective and proportionate measures to monitor and assess the general performance of authorised self-driving vehicles in Great Britain, including the extent to which that performance is consistent with the statement of safety principles.

The Secretary of State must publish a report setting out the conclusions from the monitoring and assessment each year.

Monitoring will assess the extent to which the 2 aims of the AV Act 2024 set out above are being met.

Safety principles

When assessing whether a vehicle is capable of travelling autonomously and safely the Secretary of State must have particular regard to the following considerations:

1)Self-driving vehicles should use the road in accordance with traffic laws and the Highway Code

This includes complying with road traffic laws, speed limits, signage, road markings and any temporary restrictions such as roadworks or diversions.

2) Self-driving vehicles should maintain control over the vehicle at all times.

This includes performing the driving task smoothly, safely and in a stable manner, including during unexpected events or rapidly changing conditions.

3) Self-driving vehicles should predict, detect, and respond to hazards proactively

This includes predicting and detecting potential or existing hazards, and taking safe, timely and contextually appropriate action to avoid or minimise harm.

4) Self-driving vehicles should adapt driving to road and weather conditions

This includes recognising the conditions of the road and weather and adjusting its driving strategy accordingly, to reflect the environment and maintain safe margins.

5) Self-driving vehicles should behave in a predictable manner

This includes behaving in a way other road users understand and communicating its intention through appropriate signals to other road users clearly.

6) Self-driving vehicles should improve road safety in Great Britain for all

This includes seeking to minimise risk to its occupants and other road users, avoiding any disproportionate increase in risk to any group of road users. When safety risks occur the self-driving vehicle should seek to minimise the potential for harm.

7) Self-driving vehicles should interact safely with all road users, particularly those most at risk in the event of a collision.

This includes, though is not limited to, being considerate around pedestrians, cyclists, motorcyclists, children, older adults and disabled people more at risk.

8) Self-driving vehicles should interact safely with emergency services

This includes detecting and responding safely to instructions from emergency services and other authorised officials.

9) Self-driving vehicles should only operate autonomously within the limits under which they are specifically designed to function.

This means recognising its limits (including the capability of the automated driving system to perform the driving task) and having appropriate strategies in place to respond where there is a risk that a vehicle may exceed those limits.

10) Self-driving vehicles should be designed to take account of the specificities of the territory where they will operate

This includes when considering whether a vehicle travels safely by means of comparisons to human drivers, or using data derived from human drivers, such data and comparisons should be:

representative of comparable locations and circumstances
sufficiently representative of GB drivers who are unimpaired, not distracted and follow traffic rules

In practice, responsibility for ensuring that vehicles comply with these principles will rest with the regulated body responsible for ensuring the authorised automated vehicle continues to satisfy the self-driving test.

Question 15

How to respond

Accepted Answer

We will only accept responses received before the closing date: 11:59pm on xx mmm 2026.

The easiest way to respond is to use the online response form It includes the option to save and continue your response if you are unable to complete it in one go.

If you cannot use the online form, email or post your response to us:

Email to consultation@ccav.gov.uk

Write to:

CCAV, 3rd Floor
Great Minster House
33 Horseferry Road
London, SW1P 4DR

If you send your response by email or post:

answer the questions asked in the consultation and, if required, provide further comments and evidence
tell us if you are responding as an individual or on behalf of an organisation

You will receive an acknowledgement of receipt if you submit your response by email. Check your junk/spam folder if you do not receive this within 15 minutes of sending your email. If you do not find a receipt, email consultation@ccav.gov.uk with ‘missing receipt’ in the subject line.

Question 16

Full list of questions

Accepted Answer

These questions are listed here to give you an overview of what we are asking.

Question: Do you agree or disagree with the purpose of the statement of safety principles being to provide guidance to the Secretary of State in considering the self-driving test?

Question: Why do you think this? If relevant, what alternatives would you prefer?

Question: Do you agree or disagree with the safety standard for automated vehicles being set at equivalent to careful and competent human drivers?

Question: Why do you think this? If relevant, what alternatives would you prefer?

Question: Do you agree or disagree with using behaviours as the basis for developing the principles within the statement of safety principles?

Question: Why do you think this? If relevant, what alternatives would you prefer?

Question: Do you agree or disagree with using behaviours as the basis for developing the principles within the statement of safety principles?

Question: why do you think this, and, if relevant, what alternatives would you prefer?

Question: Why do you think this?

Question: In your view, what, if any, other behaviours do you think are missing and why?

Question: Do you agree or disagree that the proposed principles on improving road safety for all and interacting safely with other road users provide sufficient protection for all road users?