As we release the latest DeepSafe research findings into public perception of automated vehicles, Behavioural Scientist, Lara Suraci explores some of the insights behind the data. What did our results tell us about the public’s understanding of road safety - and how might this inform messaging when it comes to advancing automated vehicles on UK roads?
Automated vehicles (AVs) promise more efficient travel, reduced emissions and greater accessibility for all, but what role does safety play in shaping public acceptance?
Let’s put AVs to the side for the moment and talk about road safety in the UK as it stands. In 2023, the Department for Transport puts the number of people who were killed or seriously injured (KSIs) on UK roads at a sobering 29,711, with vulnerable road users such as pedestrians, cyclists, and motorcyclists disproportionately affected.
A major contributing factor in these incidents is human error: it is estimated that 88% of road traffic accidents in the UK are caused by human error, including behaviours like speeding, distraction from mobile phones, running red lights or failing to give way.
In light of these numbers, it may come as a surprise that a survey study DG Cities conducted in December 2024 indicates that many do not consider improving road safety a priority: when we asked a representative sample of 1,000 UK respondents to select their top two priorities for the UK transport system, only 37% chose road safety; a low proportion compared to the 53% who chose affordability and the 46% who chose nationwide equality.
One possible reason for this is that the above statistics are not, in fact, common knowledge: when asked to guess the number of KSIs on UK roads in 2023, a staggering 92% of our sample reported an estimate below the official figure and the average estimate only came to 11,402 – less than half of the official figure. In other words, road safety may not be at the forefront of public concern because, quite simply, many are not aware of the scale of the problem.
How does this compare to safety expectations of AVs?
In 2024, the UK Government established the following safety benchmark for AVs: in order to be allowed to operate on UK roads, AVs should achieve an equivalent level of safety to a ‘competent and careful human driver’. Aside from the inherent ambiguity of this concept, our research suggests that public support will require AVs to not only meet but exceed this benchmark. Presented with an AV that is slightly less safe than a competent and careful human driver, a negligible 3.7% of our respondents indicated their willingness to use it; however, once the AV in question is as safe as a competent and careful human driver or slightly safer, this proportion jumps to 36.8% and 56.5%, respectively. Only once the AV is much safer than a competent and careful human driver do 3 out of 4 respondents report their willingness to use it.
Similarly, a second survey study from February 2025 revealed that the proportion of respondents who would support the widespread introduction of AVs on UK roads almost doubles when AVs promise a 10% reduction in the number of KSIs compared to no change – so safety gains over human drivers are not only a prerequisite for acceptance, but also an effective tool to boost public support of AVs.
One possible reason for this is that the above statistics are not, in fact, common knowledge: when asked to guess the number of KSIs on UK roads in 2023, a staggering 92% of our sample reported an estimate below the official figure and the average estimate only came to 11,402 – less than half of the official figure. In other words, road safety may not be at the forefront of public concern because, quite simply, many are not aware of the scale of the problem.
How does this compare to safety expectations of AVs?
In 2024, the UK Government established the following safety benchmark for AVs: in order to be allowed to operate on UK roads, AVs should achieve an equivalent level of safety to a ‘competent and careful human driver’. Aside from the inherent ambiguity of this concept, our research suggests that public support will require AVs to not only meet but exceed this benchmark. Presented with an AV that is slightly less safe than a competent and careful human driver, a negligible 3.7% of our respondents indicated their willingness to use it; however, once the AV in question is as safe as a competent and careful human driver or slightly safer, this proportion jumps to 36.8% and 56.5%, respectively. Only once the AV is much safer than a competent and careful human driver do 3 out of 4 respondents report their willingness to use it.
Similarly, a second survey study from February 2025 revealed that the proportion of respondents who would support the widespread introduction of AVs on UK roads almost doubles when AVs promise a 10% reduction in the number of KSIs compared to no change – so safety gains over human drivers are not only a prerequisite for acceptance, but also an effective tool to boost public support of AVs.
Relative Safety Expectations of AVs
How can we reconcile this deprioritisation of road safety in the general transport system with the fixation on high safety standards for AVs?
In addition to misconceptions about the current state of road safety in the UK, safety as a baseline requirement may be overlooked in everyday considerations of the transport system, but shift into focus when a disruption to the status quo – such as the introduction of AVs – is debated. Ensuring adherence to high safety standards in human drivers may also feel like a complex and unattainable goal due to the involvement of unpredictable factors like individual preferences and behaviour. Machines like AVs, on the other hand, which can be programmed to follow set rules, may offer a clearer path to safety – and thus trigger much higher expectations thereof.
But perhaps the most important question to consider is not why this discrepancy exists but how we can leverage it to foster public support for AVs and, ultimately, harness their potential for increased road safety.
While safety gains over human drivers are often cited as a key selling point for AVs – and seem to be effective in boosting public AV support – this argument may not be as powerful as it could be if the public continues to underestimate how unsafe UK roads are. In fact, the perception that road safety is not as critical a concern as statistical evidence suggests might lead to less urgency in the adoption of AVs.
Understanding and leveraging this knowledge gap could therefore play a pivotal role in advancing public support for AVs. It presents a crucial opportunity for public education and awareness campaigns that address misconceptions regarding the current state of road safety in the UK and contextualise it with regard to the true potential of AVs to make our roads safer.
Learn more
This public engagement research takes a closer look at public attitudes towards automated vehicles and expectations regarding their safety. We use approaches from behavioural economics to assess the impact of safety messaging on increasing acceptance.
Our staircase model assessed how safety statistics can be used to support acceptance of automated vehicles onto UK roads, and uncovered how different demographics engage with information on automated vehicle safety.