Skip to main content
Consultation outcome

Automated vehicles: protecting marketing terms – consultation outcome

Updated 7 July 2026

The automated vehicles: protecting marketing terms’ consultation was launched on 10 June 2025 and ran until 1 September 2025. This consultation outcome summarises the responses we received and sets out the government’s response.

It covers:

  • the background to the consultation
  • a breakdown of respondents by stakeholder group
  • the analysis methodology
  • a summary of the responses
  • government’s response
  • next steps
  • questions we received during the consultation

Background

The distinction between self-driving/automated driving systems (ADS) and advanced driver assistance systems (ADAS) is crucial. The law sets out a safety assessment process to determine whether a vehicle can safely drive itself without being controlled or monitored by a human. This process would be undermined if businesses are able to claim that their vehicles are self-driving without getting their vehicles authorised or listed.

Furthermore, drivers must understand the vehicle’s capabilities and their legal responsibilities. Marketing that describes an ADAS feature as self-driving or creates confusion about its capabilities is dangerous: it can mislead drivers into thinking that they do not need to pay attention to the road.

Misleading marketing also risks undermining public trust in self-driving technologies and fair competition with businesses whose vehicles have successfully undergone the self-driving safety assessment process.

The Automated Vehicles Act 2024 (the 2024 act) includes 2 new marketing offences to prevent businesses from misleading end-users in Great Britain into thinking that an unauthorised vehicle can safely and legally[footnote 1] drive itself:

  • The ‘protected terms’ offence (section 78) restricts the use of certain terms to authorised automated vehicles (AVs). The list of terms must be set out in secondary legislation, following consultation. It may include words, expressions, symbols or marks. This was the subject of the recent consultation.

  • The ‘confusion’ offence (section 79) is broader: it applies to any commercial communications likely to confuse end-users into thinking that an unauthorised vehicle can travel autonomously. This offence does not need secondary legislation and was not the subject of the recent consultation. However, it complements the first offence: the marketing may constitute an offence if it is likely to confuse, even if it does not use terms on the protected list.

The 2024 act will introduce a new procedure to authorise vehicles that can safely and legally drive themselves. This builds on the Automated and Electric Vehicles Act 2018 (the 2018 act) under which such vehicles must be listed. Once the authorisation procedure comes into effect, it will supersede the listing procedure under the 2018 act. The protected terms offence will only apply where the vehicle is neither listed nor authorised.

Breakdown of respondents

A total of 71 responses were received. 48 participants responded on behalf of organisations and 23 responses were received from private individuals. The breakdown of respondents is set out in Table 1.

Table 1: Breakdown of respondents by stakeholder group

Stakeholder group Number of respondents Percentage of respondents (%)
Individuals 23 32%
Safety and road user groups 9 13%
Public sector (including local government, emergency services and highways) 7 10%
Research, consultancy and academia 7 10%
Insurance 7 10%
Vehicle manufacturers and technology developers 6 8%
Legal 5 7%
Trade bodies (vehicle industry) 4 6%
Trade bodies (other) 3 4%
Total 71 100%

Individual respondents were asked if they wished to provide details of their work background. Of those who did:

  • 4 individuals were taxi drivers
  • 3 individuals worked in academia
  • 2 individuals worked in the public sector
  • 2 individuals worked in research and consultancy
  • 1 individual worked in each of the following types of organisations: emergency services, local government, legal, vehicle operator and vehicle technology developer

Analysis methodology

The consultation survey comprised a total of 28 questions and included both closed and free-text responses. All responses received were analysed by DfT officials. We are grateful to all the organisations and individuals who responded to the consultation with their views and suggestions.

Of the total of 71 responses received, 37 participants responded directly via the online survey and 34 participants responded by email. Not all participants answered all the questions. The data and analysis presented are based on the responses to each question.

Some email responses did not follow the exact question format and presented their response differently from the survey. Where appropriate, these responses have been categorised as ‘agree/disagree/don’t know’ or ‘yes/no/don’t know’ and included in the survey data. Any comments provided as part of these responses have also been considered.

Officials also used DfT’s AI consultation analysis tool (CAT) to support thematic analysis of 13 free-text responses.

See the DfT’s consultation analysis tool section for further information on how the AI CAT was used.

Summary of responses

Should terms be protected?

Table 2: Question: Conceptually, do you agree or disagree that certain terms should be protected for these vehicles?

Answer Percentage of respondents (%)
Agree 68%
Disagree 13%
Don’t know/not answered 20%

Base: all respondents (71).

Note: percentage figures in this table may not add up to 100% due to rounding.

A clear majority of all respondents agreed that certain terms should be protected for authorised self-driving vehicles.

One research organisation noted:

Yes, we are of the view that specific terms should be protected when used in relation to automated vehicles. Research evidence consistently shows that the distinction between driver assistance systems and self-driving technology is not well understood by the public. In our view, protecting certain terminology is a necessary step to ensure clear communication about the functionality and limitations of these technologies, thereby supporting informed consumer understanding and improving overall road safety.

After discussing the findings of several studies into public attitudes and trust in AVs[footnote 2], another respondent organisation concluded:

Conceptually, the idea of protecting terms and limiting their use solely to AVs should assist in alleviating the confusion regarding when a human driver can disengage from the dynamic driving task, and when they cannot. This will both drive the consistent use of terms, limit their use to AVs and help build understanding among the public as to the capabilities of vehicles.

One research organisation went further. They noted that, realistically, ‘self-driving’ capability will likely be restricted to certain conditions (the operational design domain (ODD)), such as a motorway. However, if a manufacturer says, ‘our vehicle is self-driving’, it might give an incorrect impression that the vehicle is self-driving in all conditions, including, for example, urban environments. As an end user’s mental model can be shaped by marketing communications, they recommended that protected terms should be accompanied with its high-level operating condition (ODD) description as part of the approval process check.

Those who disagreed were asked why they disagreed. The reasons given were broadly that existing laws are sufficient, that the focus should be on driver responsibility and education and that regulation stifles innovation. As one vehicle manufacturer put it:

This approach risks creating confusion, stifling innovation, and diverting attention away from where regulation matters most: ensuring clear communication and safe system design. We believe the offence already established under section 79 of the Automated Vehicles Act, together with existing consumer protection laws, provides a more effective and future-proof safeguard.

Other respondents were opposed to the development of self-driving vehicles, viewing it as a waste of resources and unsafe.

Which terms should be protected?

Table 3: Question: Do you agree or disagree that the following terms should be protected?

Terms Agree (%) Disagree (%) Don’t know/not answered (%)
Self-driving 76% 8% 15%
Drive itself 77% 7% 15%
Driverless 77% 7% 15%
Automated driving 69% 15% 15%
Autonomous driving 73% 11% 15%
Driving autonomously 75% 10% 15%

Base: all respondents (71).

Note: percentage figures in this table may not add up to 100% due to rounding.

Most respondents agreed that these terms should be protected for authorised self-driving vehicles.

Agreeing with protecting these terms

Most respondents agreed that protecting these terms was necessary to promote public understanding, prevent misuse of driver assistance systems and ensure public safety. There was strong agreement on this across different organisations, as the following comments demonstrate.

Safety and road user group

All of these terms should be protected because the consultation evidence shows they are highly likely to mislead if applied to vehicles that are not authorised as self-driving.

Advertising trade body

The consultation provides compelling evidence that these terms create consumer confusion about the level of human involvement required. The #TestingAutomation and subsequent studies demonstrate that consumers interpret these phrases as indicating full automation, making their use for driver assistance systems inherently misleading.

Local government organisation

All of the terms above give the strong impression that the vehicle can take a passenger from place to place without any need for any input from a driver.

Legal trade body

We strongly agree that the use of certain terms should be reserved for authorised automated vehicles and with the broader provision regarding the marketing of vehicles.

Vehicle industry trade body

We agree that the terms as listed in the consultation (6 specifically listed terms together with the words ‘autonomous’ and ‘automated’ when used to describe the whole vehicle) should be protected.

Vehicle technology developer

These are terms likely to be used when marketing authorised self-driving vehicles.

Respondents agreed that protected terms should only be used for a vehicle whose automation system has been officially authorised as meeting government safety standards. When a vehicle is authorised, legal responsibility for driving will transfer from the driver to the authorised self-driving entity. This was seen as a crucial distinction, which provides legal clarity in the case of an incident.

One safety and road user organisation observed:

In recent years, assisted driving systems (level 2) have been allowed to automate more and more driving tasks to assist the driver in performing dynamic vehicle control. Notably, the advanced assisted driving systems (in specific DCAS (UN R171) and especially its 01 series of amendments) may act very similarly to automated vehicles, with the main differences stemming from the role of the driver, as drivers remain ultimately responsible for the vehicle while using assisted driving systems, whereas this responsibility shifts to the system in automated vehicles.

It is, therefore, important that drivers are aware and reminded that only in automated vehicles the system becomes responsible, and that in all other vehicles (including level 2 DCAS vehicles) the responsibility remains with them. To prevent driver (mode) confusion, it is important that the listed terms may only be used in the context of systems/vehicles/situations where the automated/autonomous driving system is truly in control and responsible.

Some respondents said that if these terms are not protected, there could be a gap in the law. This is because current liability frameworks do not cover accidents caused by a driver’s misunderstanding of a vehicle’s capabilities, even if it is due to misleading marketing. There was concern that this would worsen road safety:

The work of the road safety industry risks being undermined by manufacturers making false claims about the ability of vehicles to do more than they are capable of. It is, therefore, of fundamental importance that certain terms remain protected to limit the potential for driver confusion around the capabilities of their vehicles.

Respondents thought that misleading marketing risks eroding overall public trust in automated vehicle technology, which could delay the widespread adoption of safer, more efficient transport systems. Protecting certain terms for authorised self-driving vehicles was understood as playing an important role in preventing such misleading marketing.

Several respondents stated that the regulatory framework must be flexible enough to add new protected terms in the future to adapt to evolving technology and marketing language. A few respondents also suggested that standardising terminology should be part of a wider strategy to mandate a full package of vehicle safety features.

Several respondents were of the view that, alongside protecting terms, the government should consider the role of public education and industry guidance to support the transition, to ensure compliance and to protect vulnerable road users, such as motorcyclists.

Disagreeing with protecting these terms

Of those who disagreed that these terms should be protected, some said that existing ADAS technologies already use some of these terms, and that it was impracticable to protect terms used internationally. One vehicle manufacturer said that both technology and language were undergoing rapid changes, in which terms were being used broadly:

For example, the term ‘self-driving’ has increasingly become recognised by the public as an umbrella descriptor for technologies that enhance driving safety and convenience through varying degrees of automation. While it may not carry a precise technical meaning, its broad recognition makes it a valuable linguistic tool during this early stage of public engagement with assistance systems.

Instead of protecting existing terms, respondents suggested that government should mandate the use of new, neutral and verifiable descriptors, such as a UK-specific authorised label or terms based on official standards.

Respondents also suggested that regulation should prioritise regulating the vehicle’s human-machine interface (HMI), as its design can lead directly to driver over-trust. One industry respondent argued that the focus should be on the interface, including screens and audio prompts, rather than on marketing. The respondent argued that tighter regulation of these interfaces ‘is critical, as these impact safety more directly than ads’. They suggested regulation draw on established industry safety guidelines (such as Euro NCAP) and established standards from other safety-critical sectors, such as aviation’s HMI standards.

Insurers also asked for more regulation of HMIs, but thought that this should be in addition to, rather than instead of, regulating marketing.

Different parts of speech and grammatical forms of terms

Table 4: Question: Do you agree or disagree that different parts of speech and other grammatical forms of protected terms should also be protected?

Answer Percentage of respondents (%)
Agree 65%
Disagree 4%
Don’t know/not answered 31%

Base: all respondents (71).

Most respondents agreed that different parts of speech and grammatical forms of protected terms should also be protected for authorised self-driving vehicles.

Several respondents said that protecting grammatical variations is necessary to close loopholes that marketers could exploit to circumvent regulations and mislead consumers.

They also stated that this approach aligns with public perception because the public perceives these different forms as having the same meaning as the root protected term. Therefore, affording protection to terms and their variations will provide greater clarity and enhance road safety. Respondents also said that including grammatical variations from the outset makes the regulation more concise and robust, avoiding the need to explicitly list every possible variant.

One safety organisation referred to examples from the marketing of other products:

A parallel can be drawn with restrictions on the marketing of products such as alcohol, tobacco, or high-fat, salt and sugar foods, where controls are designed to prevent terms, images or suggestions that could mislead consumers into believing the product is healthier or more appropriate than it is. The same principle should apply here: non-automated vehicles must not be marketed in a way that suggests, or could reasonably be interpreted as suggesting, that they are automated.

Some went further and suggested that new terminology within the 2024 act, specifically ‘user-in-charge’ (UiC) and ‘no-user-in-charge’ (NUiC), could become a future source of public confusion. Therefore, the communication of these new official terms should also be proactively considered.

Of the respondents who disagreed, one suggested that the language used to describe vehicle automation refers to a spectrum of capabilities rather than a clear binary state, making many terms ambiguous. Protecting variations of less precise terms could, therefore, be misleading.

Automated and autonomous

Table 5: Question: Do you agree or disagree that the following terms should be protected only when they are used to describe the whole vehicle?

Terms Agree (%) Disagree (%) Don’t know/not answered (%)
Automated 59% 23% 18%
Autonomous 63% 20% 17%

Base: all respondents (71).

Most respondents agreed that these terms should be protected when they are used to describe the whole vehicle. 

Agreeing with protecting automated and autonomous terms

Many respondents agreed that the words ‘automated’ and ‘autonomous’ are already widely used for parts of vehicle systems, such as brakes and steering, and this is not confused with a whole vehicle driving autonomously. They felt this was a balanced approach which prevented overreach while ensuring public safety and maintaining consumer protection.

As noted by a research and consultancy organisation:

The terms ‘automated’ and ‘autonomous’ should be protected only when used to describe the whole vehicle. These terms may be used in other contexts where they do not imply automated driving functionality and, therefore, are unlikely to mislead.

Industry bodies were particularly concerned that the regulations should not cover individual vehicle features or terms recommended by SAE International [footnote 3] to describe levels, such as ‘partial driving automation’:

It is also essential for the industry to be able to refer to automated functions or features in contexts that do not suggest the entire vehicle is self-driving. The term ‘automated’ applied to features or parts should remain outside the scope of these offences.

It is noted that the internationally adopted SAE J3016 definitions for terms related to driving automation systems describe the different levels of driving automation. We ask for clarification that those levels’ descriptions (for example, level 2: partial driving automation) are not within scope of the proposed protected terms provisions.

Several respondents suggested the protection should be extended slightly to also include automated and autonomous when used to describe the driving functionality or the overall capability of the vehicle. This is because language used for specific functions can still create confusion about the vehicle’s overall automation level:

This should be widened slightly to include any part of the car that could be related to the movement of the vehicle as a whole…unless the term is already widely understood to be an independent function such as automated braking.

Our recommendation: ‘automated’ or ‘autonomous’ should be considered protected when they are used to describe both driving or a vehicle.

An advertising stakeholder gave examples of boundary cases, such as ‘experience automated luxury’ or ‘the autonomous future of motoring’.

Some respondents went further: automated and autonomous should be protected when used to describe both the vehicle and its parts. It was suggested that the public should not be expected to have technical knowledge or to understand the difference between vehicle parts or features and the whole vehicle.

Disagreeing with protecting automated and autonomous terms

A significant proportion of respondents who disagreed also thought that limiting the protection given to ‘automated’ and ‘autonomous’ to whole vehicles might be too narrow.   

Some respondents said that it would be wrong to describe any ADAS steering system as automated:

We strongly recommend that equipment and parts that assist the driver in controlling the lateral and/or longitudinal motion of the vehicle on a sustained basis (such as the ACSF B1 and C categories of UN R79 as well as DCAS (UN R171)) should be covered by the protection (which would ensure ‘automated’ and ‘autonomous’ are not allowed to be used in the marketing of such assistance systems.

Other respondents accepted that it should be possible to talk about ‘automated lane keeping’ or braking but thought that confining these terms to the whole vehicle was still too narrow. It would allow a loophole where cars were advertised as providing ‘autonomous journeys’ or where an aftermarket kit was said to provide ‘automated control’.

On this basis, consultees said that applying automated and autonomous terms to the overall capability of the system should not be allowed:

In addition to ‘autonomous’ and ‘automated’ being prohibited when applied to the whole vehicle, they should also be prohibited when applied to the task of driving or to undertaking a journey by vehicle. This would, for example, prohibit references to providing ‘a journey under automated control’.

Other terms

Table 6: Question: In your view are there any other terms that should be protected under the Automated Vehicles Act 2024?

Answer Percentage of respondents (%)
Yes 41%
No 17%
Don’t know/not answered 42%

Base: all respondents (71).

Many respondents did not express a view on this question. Of those who did, most thought that other terms should be protected under the 2024 act.

Respondents were also asked what other terms should be protected and to give reasons for their view.

No other terms

Most respondents who answered ‘no’ to this question did not provide a reason for their view. Those who did said that the terms proposed give good coverage of commonly used words and phrases to describe self-driving and were largely already well received during the law commissions’ consultations. They also suggested that the list of protected terms will need to be kept under review to monitor whether any new misleading terms emerge in future.

Other terms

Among respondents who were of the view that other terms should be protected, ‘autopilot’ was the most common answer. These respondents thought that this term was misleading, as it implied that the vehicle could drive itself.

Commenting on the term ‘autopilot’, one research organisation noted:

The term is defined by the Oxford dictionary as ‘a device that keeps aircraft, spacecraft, and ships moving in a particular direction without human involvement’. Consequently, there is widespread understanding that an ‘autopilot’ refers to a system capable of automated driving without the need for human intervention or fallback control.

Another respondent observed:

This term is used colloquially in many contexts, which means that end-users may have preconceived notions of what ‘autopilot’ may indicate about a technology’s capabilities.

Further, as the National Highway Traffic Safety Administration in the US has noted, the term ‘autopilot’ can lead drivers to overestimate and over-rely on technology that does not actually have self-driving capabilities. The term ‘autopilot’ may then create exactly the sort of confusion that the law commissions wanted to avoid in their recommendations.

Some respondents went further and suggested that other iterations of the word ‘pilot’ should be included, such as ‘pro pilot’ or ‘vehicle pilot’. One respondent observed:

The term ‘pilot’ implies an entity who is in control of a vehicle, and its use in system names could lead to misinterpretation because it could be taken to mean without driver supervision. To avoid this misleading perception, we recommend that this term, and its variants, also be protected to apply only to authorised autonomous vehicles.

Respondents also suggested a variety of other terms. Some respondents thought that the word ‘robot’ or ‘robo’ should be protected, when used in conjunction with transport-related terms such as ‘driving’, ‘vehicle’ or ‘taxi’.

A few respondents were concerned with the prefix ‘auto’ when used in connection with the vehicle or the driving task, as in ‘autotaxi’ and ‘autodrive’, while a few others expressed concern over ‘hands-free’.

Other terms mentioned were level 4, level 5, chauffeur, valet parking, unmanned, driver free, ghost driving, guardian, driving intelligent car and AI car/driver.

Should only English terms be protected?

Table 7: Question: Do you agree or disagree with our approach of only protecting English terms?

Answer Percentage of respondents (%)
Agree 41%
Disagree 21%
Don’t know/not answered 38%

Base: all respondents (71).

Many respondents did not express a view on this question. Of those who did, a majority agreed with our approach of only protecting English language terms.

Those who disagreed were asked why they disagreed. Most were concerned that foreign words could be used to circumvent the protection:

Some foreign language words are well understood and might be used to market vehicles originating from those countries. There could be foreign brands coming with their own brand name/marketing term for a vehicle or a service that is just slightly different to the English word but not protected.

Examples given were the German ‘autonomes fahren’ and French ‘conduite autonome’. An independent safety and road user group said that ‘government should also look to protect terms in Welsh’.

Some consultees thought that it would be impractical to provide a full list of foreign terms. Instead, they suggested the problem should be addressed in other ways:

It would be better addressed through further clarification of what constitutes another grammatical form of a word, as opposed to attempting to list all the possible non-English words, which would be impractical.

Whilst it may not be within the control of DfT to protect non-English terms, a good practice guide to terms can be suggested by DfT.

Table 8: Question: Do you agree or disagree that there will be sufficient legal safeguards to prevent the protection from being applied to marketing which is unconnected with driving automation?

Answer Percentage of respondents (%)
Agree 32%
Disagree 10%
Don’t know/not answered 58%

Base: all respondents (71).

Many respondents did not express a view on this question. Most of those who did agreed that there will be sufficient legal safeguards to prevent the protection from being applied to marketing unconnected with driving automation.

Respondents were asked to give reasons for their answer. Those who agreed generally said that the safeguards were reasonable and sufficient. Some respondents also cited section 79(4) of the 2024 act:

For example, section 78(4) of the 2024 act applies where the accused proves that the protected term was ‘used in a way which was not intended to convey and could not reasonably be understood as conveying, any meaning to do with automation’.

Among those who disagreed, the most common theme was the need for greater clarity for businesses. This view was also shared by some of those who agreed that there would be sufficient legal safeguards. These respondents asked for guidance on this issue to provide more clarity.

Symbols and marks

Table 9: Question: In your view are there any specific symbols and marks that indicate a vehicle is self-driving that deserve special protection?

Answer Percentage of respondents (%)
Yes 8%
No 37%
Don’t know/not answered 55%

Base: all respondents (71).

Many respondents did not express a view on this question. Of those who did, most were not aware of any specific symbols or marks that indicate a vehicle is self-driving that deserve special protection under the 2024 act. Some respondents felt it was too early to protect a specific symbol or mark at this stage.

Those who answered yes to this question were asked to specify which symbols or marks they thought deserved special protection and to give their reasons.

Only 3 respondents specified symbols which, in their view, deserve special protection and each gave a different answer. One respondent considered that if terms like SAE L4 or L5 are protected, then the associated marks should also be protected. Another respondent said that, while they thought SAE symbols were a good representation of the person in the driver’s seat of a semi/ fully autonomous vehicle, they thought these symbols ‘would need to be slightly enhanced’. They did not provide further details.

A third respondent referred to the cyan automated driving systems marker lamp (ADS ML) being developed by a UNECE task force on the topic (TF AVSR of the GRE at WP 29).

Even though so few respondents indicated a current symbol, respondents generally saw merit in having a recognisable and protected symbol to indicate a vehicle is self-driving. One legal respondent observed:

Just as wi-fi has become universally denoted… it would be beneficial for AVs to have a universally recognised symbol.

It was noted that some connectivity-related symbols are currently being used for driver assistance systems. These symbols could confuse the public by suggesting that the vehicle has self-driving capability. One way of addressing the problem would be by having a recognised symbol specifically for self-driving vehicles.

Some respondents suggested that government should take a lead in developing a nationally recognised symbol or kitemark with stakeholder buy-in. Others said that government should pay close attention to symbols used by vehicle manufacturers to denote self-driving and seek to protect those symbols that become widely recognised.

Business costs

Table 10: Question: If the terms proposed in this consultation were protected, would your business incur any costs as a result?

Answer Percentage of respondents (%)
Yes 20%
No 70%
Don’t know/not answered 10%

Base: 10 organisations that answered this question.

This question was asked of respondents who indicated that they were responding on behalf of an organisation involved in the vehicle industry, either as a manufacturer, supplier, operator or developer or as a trade body. Most respondents to this question advised that their business would not incur any costs if the terms proposed in this consultation were protected.

Those who indicated their business would incur costs were asked to provide details about the scale of costs and how these costs were determined. Some respondents felt it was too early to provide a reliable estimate of the costs.

One vehicle manufacturer did not address the cost question directly. However, they commented more generally that for global businesses, there were significant operational challenges in maintaining multiple versions of product names, marketing content and consumer-facing materials that ‘imposes disproportionate cost and complexity’.

Other respondents said that the cost for businesses is likely to be minimal, one-off costs to comply with the new regulation, such as costs associated with updating marketing materials, legal documents and training staff. It was recognised that these costs would only apply to businesses offering products to end-users and could be minimised by setting out the protected terms clearly.

Responding on behalf of its members, one industry body stated:

…members involved in research projects that involve the deployment of automated vehicles, conclude there will be some effort required to comply, but it should be fairly minimal as long as there is clear guidance and terms are correctly defined.

This view was shared by insurers who anticipated costs associated with the need to clarify policy wordings to ensure accurate and effective coverage. At the same time, they also recognised the benefits these changes would bring.

One insurance organisation noted:

The industry is supportive of these changes and firmly believes that the benefits— namely, increased clarity for consumers, improved road safety, and enhanced trust in automated vehicle technology — will ultimately outweigh the associated costs.

Insurers also pointed to the costs of not preventing misleading marketing. This ‘could result in drivers placing undue reliance on automation, leading to a higher risk of accidents and, consequently, increased claims costs for insurers’.

Other comments

Respondents were provided with an opportunity to make additional comments, and many did so. Most of the comments referred to points already made in response to other questions in the consultation. Some respondents took the opportunity to thank the government for raising this important issue.

Several respondents took the opportunity to ask for clarification on various points. See the questions asked during the consultation section for more information.

Two additional points emerged:

  • the importance of a wider programme of education, research and review
  • setting standards or guidance for ADAS technologies

A wider programme of education, research and review

Many consultees argued that the list of protected terms must be kept under review through a programme of research. It should also be combined with wider education about self-driving technology:

It is, therefore, important that the department/CCAV keep the market under review and be prepared to swiftly add further terms to the protected list should problematic brand names emerge and to include misleading non-English terms where necessary.

Government should set out a process for protecting additional terms in future. It is not possible to anticipate all future terminology, so the legislation must be adaptable.

Several consultees, including insurers, stressed the importance of public education about what AVs can (and cannot) do. One safety and road user group highlighted that education campaigns have a ‘proven track record’, for example, through the THINK! campaigns.

Another safety group considered it vital that clear, accessible information should be made available through campaigns and educational materials.

It was acknowledged that public education and awareness required collaboration between stakeholders:

Industry, government, and consumer agencies must work together to inform and educate the public, addressing misconceptions and promoting safe, responsible use.

ADAS technologies

Several consultees suggested that government should also take a more proactive approach to monitor the way that ADAS systems are marketed:

Government should set out how the public, local authorities, media and other stakeholders will be informed about the correct use of terms relating to AVs and those relating to ADAS.

Given that the autonomous vehicle environment will develop at pace, we propose that marketing for each new vehicle should be assessed before approval.

Consultees were generally concerned about the distinction between level 2 and level 3 systems and that marketing of level 2 systems, which require the driver to be in full control of the vehicle, can create a perception of greater automation than exists. There was concern that this encourages over-reliance on the system and disengagement from the driving task, which risks public safety.

Citing evidence of increased driver inattention with hands-on driver assistance systems as compared to manual driving, and that this driver inattention is exacerbated with hands-off systems[footnote 4], one safety and road user organisation concluded:

The research suggests that driver(s) want to pass more responsibility to the vehicle than it is capable of. It is in this environment of varying levels of responsibility[footnote 5] that the protected terms should come to be used purely when complete responsibility has been handed over to the vehicle and its operating system. To use them in any other use would lead to user confusion.

Another safety organisation observed:

Companies should seek to minimise to the greatest extent possible any opportunity for misrepresentation.

Some consultees recommended that there should be guidance for industry on how to communicate ADAS features. Respondents suggested this could be provided by third-party organisations:

Accompanying guidance on best practice for communicating non-self-driving features to help avoid confusion would be a beneficial resource to enable the implementation of the policy – providing those affected by the policy with additional clarity. This recommendation is just guidance and not to be included in the policy and may be provided by third-party organisations.

This was accompanied by calls to standardise the HMI of ADS and ADAS systems:

We believe that it may also be beneficial to standardise a tell-tale or consider other appropriate messaging methods for informing the user whether the automated system is engaged. It should be clear to drivers whether the ADS is currently active…to avoid ambiguity and further support road safety. Consumers should not be exposed to a variation of indicators across different vehicles, as this will increase confusion.

Misuse in in-vehicle HMIs (for example, labelling and instructing with physical labels or audio commands, whether AI-generated or not) is already leading to over-trust, as seen in incidents where drivers disengage due to confusing interfaces is even more important to be regulated.

Government response

The responses to the consultation showed strong support for protecting certain marketing terms so that they can only be used in connection with vehicles that are authorised or listed as automated vehicles, meaning they can drive themselves safely and legally.

Following consideration of the responses to the consultation and the other correspondence we received, government has decided to bring forward a statutory instrument to protect terms under section 78(1) of the 2024 act.

Government has already introduced a procedure to list vehicles as capable of safely driving themselves under section 1 of the 2018 act. A vehicle listed under the 2018 act will be treated as an ‘authorised automated vehicle’ for purposes of the 2024 act’s marketing provisions until the safety authorisation procedure under the 2024 act is brought into effect.[footnote 6]

These safety procedures would be undermined if businesses can describe their vehicles using terms that indicate they are self-driving, even though the vehicles are neither listed nor authorised as capable of safely driving themselves.

The distinction between self-driving/ADS and ADAS is crucial for clear public understanding of each technology and its safe use. At present, drivers are confused about this distinction, which has the potential to contribute to overreliance on systems which are not self-driving, causing a danger to road users. The problem is aggravated if marketing gives drivers the misleading impression that they do not need to monitor the road while driving.

This presents clear public safety risks and risks undermining the gains that might be expected from improvements in driving automation technology. This is a problem now and, if not addressed, will become a bigger problem when genuinely self-driving vehicles become available.

The public also need to trust this emerging technology. Marketing claims which indicate vehicles are self-driving when they are not undermine this trust. They also undermine fair competition with businesses whose vehicles have successfully undergone the self-driving safety authorisation or listing process. 

Government has, therefore, decided to make the statutory instrument. This is for 4 main reasons:  

  • to safeguard the integrity of the safety authorisation scheme under section 3 of the 2024 act

  • to promote fair competition, and encourage investment in the development of authorised automated vehicles, by preventing vehicles that are not authorised from being marketed as if they were

  • to protect public safety by mitigating confusion and ignorance among drivers surrounding the capabilities of vehicles that are not authorised automated vehicles and the responsibilities of drivers operating them

  • to build trust in self-driving technology

Disagreement with protecting marketing terms

There was some disagreement. It was claimed that ‘self-driving’, for example, is an umbrella term, used globally to describe varying degrees of automation. It was said that if emerging ADAS technologies could not be described in this way, new technologies would be less likely to be adopted, preventing the realisation of their safety benefits and stifling innovation.

Furthermore, it was said that any confusion could be addressed in other ways, including through information at the point of sale, system design and prosecutions under the section 79 confusion offence of the 2024 act.

One interested party submitted detailed observations following the closure of the consultation. They contended that the evidence cited in the consultation was limited, outdated and methodologically deficient, and cannot be relied upon as a basis for proceeding with the regulations. It was said that the evidence does not establish a causal link between specific marketing terms that would be protected and consumer confusion, and that the evidence points to consumer education, not marketing term restrictions, as the appropriate remedy.

It was claimed that key findings from the evidence includes a low overestimation rate that indicates that the scale of genuine dangerous overestimation is very small.

We have considered these arguments carefully and have decided to proceed with the regulations. Our reasons are summarised below. The specific comments on individual studies are summarised and discussed in the de minimis assessment published alongside the regulations.

Meaning of ‘self-driving’

We do not accept that self-driving is an umbrella term covering a variety of driver assistance technologies. Under the 2024 act, a vehicle only meets the self-driving test if it does not need to be monitored or controlled by an individual. Under section 1, it must be able to travel safely and legally by means of the vehicle’s systems without human input and when ‘neither the vehicle nor its surroundings are being monitored by an individual with a view to immediate intervention in the driving of the vehicle’. 

The test under the 2024 act is judged to be in line with a general global consensus. A recent analysis of this issue by a leading US academic concluded that there is a broad consensus among developers and regulators of motor vehicle technologies ‘that the term self-driving correctly refers only to a system whose user does not need to pay attention’. [footnote 7]

There are many existing and new technologies designed to assist rather than replace the driver, but which still require the driver to pay attention to the road and always remain in full control of the vehicle. Government agrees that these advanced driver assistance systems have the potential to make driving safer and that their uptake is desirable to promote road safety.

Government is not restricting the use of ADAS technologies, only the use of certain terms to describe them. Government considers that the use of terms to market ADAS that oversell or mislead as to their capabilities undermines trust in both ADAS and AVs, inhibiting the development and uptake of these beneficial technologies.

Confusion offence

We have considered whether it would be possible to rely on the confusion offence in section 79, without having a list of protected terms. The benefit of having a legally defined list of terms under section 78 is that it provides clarity for industry and other stakeholders. Moreover, the 2024 act intends that both offences should work together and complement each other to effectively regulate misleading marketing.

Government does not believe that relying solely on the confusion offence will achieve, or sufficiently advance, the objective of preventing businesses from using these terms in their marketing of driving automation. Government considers that only implementing the protected terms offence alongside the confusion offence will facilitate clear and effective regulation of misleading marketing, and that this is necessary to meet the four policy objectives outlined above.

Evidence of confusion

There is a consistent body of evidence to show that the public is confused about the dividing line between ADAS systems and self-driving. The consultation cited several studies and reports between 2018 and 2023. DfT published further studies following the launch of the consultation. [footnote 8] [footnote 9] [footnote 10] These studies and reports have consistently shown high levels of misunderstanding and confusion in GB over how drivers should engage with these new driving automation technologies. This is a source of serious risk to the public, which cannot be ignored.

While there may be limitations to any one study, the body of evidence showing driver confusion and misunderstanding in Great Britain is clear and consistent across multiple studies, conducted by different well-established organisations over time.

Ipsos report

DfT published an independent study by Ipsos on public understanding of vehicle automation terms. Consistent with the body of evidence and the studies cited in the consultation, the report found that people in GB ‘had limited understanding regarding the capabilities and legal responsibility for different car types and ADAS/ADS features and brands’.

The Ipsos report confirmed that there is continuing, widespread confusion over ADAS terms and capabilities, with the people most confident in their knowledge being those most likely to overestimate the capability of the system. This combination of confidence and overestimation is particularly dangerous, as these drivers may be more inclined to disregard the various warnings they are given about the need to pay attention.

The Ipsos report was published while the policy has yet to be finalised. DfT invited comments and observations on the report and was open to receiving observations from all interested parties in relation to the research and in relation to the proposed regulations in light of that research. 

Following publication, observations were received from one interested party on what was characterised as the late disclosure of the report, its methodological robustness and the soundness of any conclusions government may seek to draw from it. These observations were carefully considered and are summarised and discussed in the de minimis assessment, published alongside the regulations.

Similar to observations made in relation to the evidence cited in the consultation, it was contended that the Ipsos research contained serious methodological flaws that render its findings unreliable as a basis for the regulations.

Government does not accept this contention. The Ipsos report is a recent example of a significant body of work that shows that the public is confused about ADAS technologies and does not understand the crucial distinction between vehicles that meet the self-driving test (and must be listed or, in future, authorised as capable of driving themselves safely and legally) and those that do not. This is a problem now and, if not addressed, will become a bigger problem when genuinely self-driving vehicles become available. Government cannot ignore the risks this presents for public safety.

Combatting driver confusion and the role of marketing

When an ADAS feature is engaged, it is crucial for road safety that drivers have a clear understanding of: 

  • the capabilities and limitations of the ADAS feature

  • the need to actively monitor the road and always maintain full control of the vehicle

  • their legal responsibilities as drivers

In March 2026, DfT published an evidence review by the National Centre for Social Research (NatCen) on understanding human factors in advanced driver assistance systems. The report presents findings from 59 academic literature studies from 2017 onwards and provides a robust overview of the existing evidence regarding the role of human factors in ADAS technologies.

Consistent with the body of evidence, the Ipsos report and the studies cited in the consultation, a key finding of the review of the existing literature was that ‘there was limited understanding among drivers of their responsibilities and capabilities and the operational boundaries of ADAS’. [footnote 11]

The review demonstrated from the existing literature that when using Level 2 ADAS, drivers experienced challenges related to understanding their responsibilities and the system’s capabilities.[footnote 12]

The authors conclude: 

The policy implications from this review suggest a multifaceted approach that includes enhanced driver education, improved system design, robust data protection, and comprehensive research to ensure the safe and effective use of Level 2 ADAS technologies. Findings underscore the necessity for enhanced communication and education about their limitations and proper use. Policymakers should consider creating comprehensive, evidence-based educational programmes that inform drivers about their responsibilities and the system’s assistive nature.

As part of this effort, the authors recommend engaging with automotive manufacturers ‘to promote clearer and more accurate communication regarding ADAS capabilities and driver responsibilities in their marketing and advertising materials’. [footnote 13]

Combatting driver confusion requires actions by industry and government. Education at the point of sale is clearly important – though it only applies to buyers, not all drivers. In-vehicle measures such as warnings, driver attention monitoring systems and human interface design all have a part to play. Clear and accurate marketing also has an important role.

Following publication of the evidence review, some observations were received from one interested party, similar to those received in relation to the evidence cited above. These observations were carefully considered and are summarised and discussed in the de minimis assessment, published alongside the regulations.

It was contended that the evidence review was limited, there are no studies that suggest a marketing ban is required, and the remedies proposed are driver education, driver monitoring systems and multi-modal feedback. Therefore, it cannot be relied on as a basis for the regulations.

While there are some limitations to the literature review as a rapid evidence assessment (which are set out in the evidence review itself), the conclusions set out in the evidence review were found to be established notwithstanding the limitations identified.  

The review’s findings and recommendations support our conclusion that regulating misleading marketing is a justified and proportionate part of a broader strategy to combat driver confusion over ADAS. Government is satisfied that the existing body of evidence, read as a whole and including the Ipsos report and the studies cited in the consultation, provides a sufficient basis for the intervention. 

Government is not implementing a marketing ban on ADAS technologies, only restricting the use of certain terms to describe them. Providing accurate information to drivers regarding system capabilities has been shown to improve trust and public acceptance.[footnote 14]

Basis for making regulations

The various contentions regarding the evidence base only address the issue of driver confusion and whether the evidence of driver confusion among GB drivers today can be relied upon to justify regulatory intervention. They do not address the other policy objectives for protecting certain terms only for authorised self-driving vehicles. These are to:

  • safeguard the integrity of the listing/authorisation process
  • ensure fair competition between those manufacturers whose vehicles meet the requirements for listing/authorisation and those that do not
  • promote trust in driving automation technology

Even on the issue of driver confusion, the basis for government making regulations is not solely on published research. Wider factors were also considered. The consultation cited the law commissions’ 4-year review, which specifically recommended protecting 5 of the 8 terms. The consultation mentioned that 3 of these terms are used in legislation to refer to vehicles where human monitoring and control are not required, at least in some circumstances.

The consultation also cited evidence of market failures, specifically in relation to public safety and the asymmetry of information between manufacturers and consumers. The consultation provided a high-profile example of an accident in China, which resulted in the tragic loss of life and cited a report from the Centre for Data Ethics and Innovation (CDEI) that noted ‘there is currently public confusion, exacerbated by some claims from industry, about the capabilities and limits of AV systems’.

The CDEI report also noted that, in the absence of regulation, companies are incentivised to downplay the operating conditions (called the operational design domain (ODD)) under which a driving automation system or feature is specifically designed to function in public communications ‘and overstate the capacity of their technology’. When it comes to questions of liability, ‘companies will have the opposite incentives, to claim that their ODDs are narrow’. [footnote 15]

Finally, we note the strong support for protecting marketing terms in the consultation responses.

Benefits and costs

It was suggested by one respondent that the intended regulations would create significant operational challenges and impose disproportionate cost and complexity in updating product names, marketing content and consumer-facing materials for GB compared to those for other countries.

Government has carefully considered the representations made in relation to business costs and carried out a de minimis assessment that is published alongside the regulations.

To the extent that the regulations would entail costs to businesses operating in the UK and across other jurisdictions in the short-term, government considers this burden to be proportionate and justified. It is outweighed by the benefits to public safety and the benefits to businesses of fair competition and promoting public understanding and trust in the long-term.  

Protecting certain terms for authorised self-driving vehicles will provide businesses with clear parameters around how to market driving automation technologies to consumers, which mitigates against the risk of confusion, misunderstanding or over-reliance. This, in turn, will help build public trust and confidence in both driver assistance technologies and true self-driving vehicles and drive future adoption, innovation and growth.

The main terms 

There was strong agreement that the following terms should be protected: 

  • self-driving 

  • drive itself 

  • driverless 

  • automated driving 

  • autonomous driving 

  • driving autonomously 

This is supported by independent research into public understanding of these terms.[footnote 16] There was also strong agreement that different parts of speech and other grammatical forms of these terms should also be protected. 

Government has now specified that these terms, different parts of speech and other grammatical forms of these terms, are protected terms under section 78(1) of the 2024 act. These terms may only be used to describe vehicles that are authorised or listed as able to safely and legally drive themselves. These terms must not be used to describe other vehicles, including those fitted with driver assistance systems.

Automated and autonomous

The consultation proposed that the terms automated and autonomous should be protected, but only when they are used to describe the whole vehicle and not when used to describe specific vehicle parts or vehicle equipment.

On this basis, it would be lawful to market an unauthorised vehicle as having automated windscreen wipers or autonomous emergency braking. However, it would be unlawful to describe it as an ‘automated car’ or an ‘autonomous vehicle’.

Having considered the consultation responses, we accept the view expressed by industry stakeholders that it should be permissible to market individual vehicle features as automated or autonomous in contexts that do not suggest the whole vehicle is self-driving.

However, we also agree with those respondents who argued that it would be too narrow to protect these terms only when used to describe the vehicle. This could simply encourage creative marketing, for example, where a car was advertised as providing ‘autonomous journeys’ or where an aftermarket kit was said to provide ‘automated control’.

Therefore, we have made a small change to regulation 2(2) of the statutory instrument since the draft version published alongside the consultation. The terms automated and autonomous will be protected when used to describe a whole vehicle and when they are used to describe the ‘overall driving functionality or capability of a vehicle’.

Other terms

When asked about other terms that should be protected, consultees gave a long list of potential terms, including ‘pilot’, ‘robotaxi’, ‘chauffeur’ and ‘AI driver’.

We recognise that there may be other terms used in the future, which could give a misleading impression that a vehicle is able to drive itself. This is why the protected terms offence under section 78 of the 2024 act operates alongside the confusion offence in section 79.

Where a business communication would be likely to confuse end-users into thinking that an unauthorised vehicle can travel autonomously, safely and legally on roads and other public places in GB, government intends to take appropriate action under section 79.

For now, government has not specified any other terms under section 78(1). However, we will keep this issue under review, with the possibility of protecting further terms in future.

Non-English terms

Government has decided to only protect terms in the English language. If terms in other languages are used in future to circumvent the protection and in a way which would likely confuse end-users, government reserves the right to take appropriate action under section 79.

Government remains open to the possibility of protecting non-English terms in the future.

Symbols and marks

Government has decided not to protect any symbols or marks at this stage. It seems clear that, at present, there is no specific symbol or mark that is commonly understood to indicate a vehicle is self-driving.

A specific symbol or mark might emerge in future as the industry develops. If so, it could fall within scope of the section 79 offence if it is used in a business communication and used in a way that is likely to confuse end-users into thinking that an unauthorised vehicle can drive itself safely and legally. In such an event, government reserves the right to take appropriate action under section 79.

Government remains open to the possibility of protecting specific symbols in future.

Government believes that the 2024 act provides sufficient legal safeguards to prevent the protection from being applied to marketing unconnected with driving automation.  

Section 78 sets out three specific legal defences. Under section 78(4), a protected term can be used in connection with a road vehicle if it ‘was used in a way that was not intended to convey, and could not reasonably have been understood as conveying, any meaning to do with automation’. For example, where an advert for ‘self-drive van hire’ would be clearly understood to mean that a human hirer is expected to drive the vehicle, it would not be an offence under section 78. 

Section 78(5) provides a defence where the term is directed only at end-users or potential end-users outside Great Britain. The business has a defence if it took all reasonable precautions and exercised all due diligence either to prevent the term from coming to the attention of end-users in GB, or to ensure that end-users in GB would understand that the term was not directed at them.  

Section 78(6) protects those not involved in the manufacture or supply of a vehicle, when the person did not know and had no reason to suspect that the use of the term would amount to an offence,   

Given these legal protections, the government is satisfied that the regulations are proportionate and strike the right balance between promoting public safety and protecting legitimate commercial activity. 

However, the offence is new. Many respondents asked for further clarification about how section 78 will work in practice. See the questions received during the consultation section for more information.

Other comments

Wider programme of education, research and review

Government notes the responses provided and will consider how best to update the list of protected terms in future in response to evolving language and marketing. 

Government will also consider the role of public education and awareness and how this can be supported by stakeholders such as government, industry and third-party organisations.

ADAS technologies

Government notes the comments and concerns expressed in relation to ADAS technologies and will consider these as part of DfT’s wider work on ADAS policy. Government continues to work internationally to create harmonised regulations for ADAS technologies and considers this the most appropriate approach.

As set out in the Road Safety Strategy, government will continue to engage with the automotive sector to ensure ADAS technologies are fully utilised and to understand any barriers to adoption.

Next steps

Government laid the statutory instrument the Automated Vehicles (Marketing Restrictions) Regulations 2026 in Parliament on 7 July 2026. The regulations will be subject to the negative procedure and are expected to come into force on 7 January 2027. 

This responds to representations made by industry that they will need time to adapt to the changes.

Questions received during the consultation

Given that the marketing provisions include some novel concepts, below we seek to address the points regarding which consultees had asked for clarification during the consultation. It is acknowledged that interpreting legislation is ultimately the role of the courts.

SAE levels of driving automation

Are the level descriptions under the SAE International’s J3016 definitions for terms related to driving automation systems (for example, level 2: partial driving automation) within scope of the protected terms provisions?

No, these level descriptions are not within scope of the protected terms. However, as the consultation explained, using words such as ‘partial’ to caveat any protected term, for example, partial ‘self-driving’, in connection with a vehicle that is not authorised or listed would fall within scope of the provisions.

Organisations already using protected terms

Some organisations may already be using protected terms. Will they have to change them?

Not necessarily. Any organisation already using a protected term for their vehicle technology has the option to get their vehicle listed under existing legislation providing it can satisfy the relevant criteria.

The consultation explained that vehicles that are capable of safely driving themselves must be listed under section 1 of the Automated and Electric Vehicles Act 2018. At present, listed vehicles will be treated as authorised vehicles for misleading marketing (2024 act, s 81(5)). However, once the authorisation procedure under the 2024 act comes into force, it will supersede the listing procedure under the 2018 act.

Therefore, under the current law, only listed vehicles will be allowed to be marketed using the protected terms. In future, the intention is that only authorised vehicles will be allowed to be marketed using protected terms. The protected terms offence will only apply where the vehicle is neither listed nor authorised. An organisation already using a protected term will not be able to continue doing so in relation to a vehicle that is not listed.

Enforcement

We would welcome clarity on the envisaged enforcement framework for the new marketing restrictions. Who would be the relevant enforcement authority?

Schedule 5 of the act sets out a framework for the civil enforcement of the marketing provisions. It is envisaged that most enforcement will be conducted through this framework. The act places a duty on the Secretary of State for Transport to enforce the marketing offences. 

In practice, we expect enforcement may be carried out by one of DfT’s agencies. The designated enforcement authority will likely issue guidance on how they intend to enforce the marketing offences. The enforcement authority may also conduct market surveillance as part of its enforcement role.

Meaning of ‘end-user’

A company involved in a trial deployment marks its vehicles as ‘self-driving’ and a transport operator or its employee is misled into believing the vehicle is authorised/ listed when in fact it is not. Could the transport operator or its employee be considered an end-user?

The protected terms offence only applies if the communication is ‘directed at an end-user or potential end-user of the vehicle’. The term ‘end-user’ is defined in section 81(1) of the act as ‘a person who uses the vehicle on a road or other public place’ for purposes which are not to do with ‘the development, manufacture or supply of the vehicle’.

The consultation explained that passengers in an automated bus or taxi-type service are unlikely to be considered ‘end-users’ as they do not control, manage or operate the vehicle. Therefore, a trial of an automated passenger service could be described as ‘self-driving’, even if it had not been authorised and relied on a safety driver. This is because the term would be directed at passengers rather than drivers or operators. The only person ‘using’ the vehicle is the safety driver, who does so as part of the development of the vehicle.

However, once the trial has ended and the vehicle is yet to be authorised or listed, if the vehicle company continued to mark its vehicles as ‘self-driving’ to end-users, then this would fall within the offence. An employee of a transport operator would likely be considered an ‘end-user’. A ‘person’ can also be a company.

On this basis, it is possible that the transport operator may also be considered an ‘end-user’. Once the vehicle has been listed/ authorised, it would be permissible for the vehicle to be marked as ‘self-driving’.

Communications with investors and business partners

How will end-user marketing be distinguished from non-end-user marketing, including technical collaboration and communications with investors and business partners? How can we be sure that legitimate marketing not directed at the end-user does not fall within the scope of the proposed offence?

The consultation explained that the offences are not intended to limit academic discussion, nor do they apply to communications between investors and developers. As mentioned above, the act defines an ‘end-user’ as ‘a person who uses the vehicle on a road or other public place’. Investors and business partners involved in technical collaboration and do not ‘use’ the vehicle on a road or in a public place and are unlikely to be regarded as ‘end-users’.

Therefore, communications directed at investors and at business partners for technical collaboration will likely fall outside the scope of the offence.

Marketing in foreign jurisdictions

A company has deployed driverless passenger services in a foreign jurisdiction. It may be considering deploying such services in GB and will likely be using vehicles virtually identical to those deployed in foreign jurisdictions. It will already be marketing its driverless passenger services in those foreign jurisdictions. What are the implications of that marketing coming to the attention of individuals in GB where that had not been the original intention?

Under section 78, it may be an offence to direct protected terms at end-users from another country, if it is ‘reasonable to anticipate’ that the term will come to the attention of end-users in Great Britain. However, the business has a defence if it can prove that the protected term was only directed at end-users outside GB.

Also, that it had taken ‘all reasonable precautions and exercised all due diligence’ to prevent the protected term coming to the attention of end-users in GB or, if it did come to their attention, to ensure that end-users here understand that the term was not directed at them.

The consultation gave an example of a manufacturer who sets up a website which may be accessed in GB, claiming that its vehicle has been authorised to drive itself in California. This would not be an offence provided the manufacturer can show that its marketing was not directed at end-users in GB and it had taken all reasonable precautions and exercised all due diligence to ensure that end-users would understand that the vehicle cannot drive itself on roads in GB.

For the confusion offence, a similar defence applies under section 79(3).

Secondary involvement

Does the defence for secondary businesses, which are not involved in the manufacture or supply of the vehicle or equipment in question, apply to media providers other than newspapers?

Under section 78(6), businesses which are not involved in the manufacture or supply of the vehicle or equipment in question have a defence if they can show that the communication that used a protected term was in the course of another business and that they did not know and had no reason to suspect that the communication would constitute an offence.

The consultation provided an example of a newspaper carrying an advert wrongly describing a new car model as ‘self-driving’. At first sight, the newspaper has ‘caused or permitted’ an illegal communication. However, it has a defence if it has no reason to suspect that the vehicle was not authorised or listed.

Once the regulator has notified the newspaper, the newspaper can no longer use this defence if it continues to carry the advert. This defence is not limited to the newspaper business. It would also apply, for example, to a website or podcast that carried the advert.

For the confusion offence, a similar defence applies under section 79(4).

DfT’s consultation analysis tool (CAT)

The AI consultation analysis tool (CAT) was used to extract themes from free-text responses and, following human review by the policy team, classify (map) human-validated themes to responses and produce summary statistics (for example, estimated number of theme mentions).

The CAT was developed by the DfT AI and data science team in collaboration with the Alan Turing Institute. Details on the CAT methodology and evaluation results can be found in the AI consultation analysis tool evaluation

Using the CAT resulted in a dataset with 13 analysed free-text questions. The results were used to inform the consultation response.

All questions were answered voluntarily, and respondents were able to ‘skip’ questions throughout the survey. Therefore, the total number of responses (to both closed and open questions) varies between questions. 

  1. Driving ’safely’ and ’legally’ in this context is a reference to the self-driving test in section 1 of the Automated Vehicles Act 2024

  2. Tennant et al, driverless futures? A survey of UK public attitudes, May 2022. Yokoi, trust in self-driving vehicles is lower than in human drivers when both drive almost perfectly, Transportation Research Part F: Traffic Psychology and Behaviour, Vol 103, May 2024, pp 1 to 17. Zhang et al, public perception of autonomous vehicle capability determines judgment of blame and trust in road traffic accidents, transportation research part A: policy and practice, vol 179, Jan 2024. 

  3. Taxonomy and definitions for terms related to driving automation systems for on-road motor vehicles

  4. Carsten, O and others (2023), Driver attentiveness to the driving task during ADAS use (PDF)

  5. For clarification, there is no variance in levels of responsibility for ADAS – the driver is always responsible and must always remain in full control of the vehicle. Responsibility for the driving task only shifts for an authorised self-driving vehicle under the 2024 act. 

  6. See section 81(5) of the 2024 act. 

  7. Smith, Bryant Walker, self-driving means self-driving (August 01, 2025). Forthcoming in Drake Law Review. See alternative website to access the publication

  8. Technology Tracker: Spring Wave 2025, (PDF) August 2025. 

  9. Understanding human factors in advanced driver assistance systems: an evidence review March 2026. 

  10. Public understanding of vehicle automation terms March 2026. 

  11. Understanding human factors in advanced driver assistance systems: an evidence review, March 2026, page 5. 

  12. Understanding human factors in advanced driver assistance systems: an evidence review, March 2026, page 47. 

  13. Understanding human factors in advanced driver assistance systems: an evidence review, March 2026, page 49. 

  14. Self-driving technologies need the help of the public, PAVE UK, May 2025. S. Khastgir, S. Birrell, G. Dhadyalla, and P. Jennings: Calibrating trust through knowledge: Introducing the concept of informed safety for automation in vehicles, Transp Res Part C Emerg Technol, vol. 96, pp. 290 to 303, 2018. 

  15. Responsible innovation in self-driving vehiclesCDEI, August 2022. 

  16. A survey conducted by Ipsos in 2025 of a representative sample of 2,186 individuals across the UK found that around half or more than half of people thought a ‘self-driving’ car (57%), a car that ‘drives itself’ (57%), a ‘driverless’ car (58%), an ‘autonomous car’ (49%) and a car that has ‘automated driving’ (48%) will be able to drive themselves with no attention or only some attention from a human. What does it mean? Public understanding of vehicle automation terms 2026, section 6.1.