Public opinions and social trends, Great Britain: August 2025

The following information is from data collected from 6 to 31 August 2025, based on adults in Great Britain.

• When asked about the important issues facing the UK today, the most commonly reported issues were the cost of living (87%), the NHS (82%), and the economy (70%).

• When asked whether the use of artificial intelligence (AI) would benefit them, 41% of adults strongly agreed or agreed; this proportion has remained broadly stable since the question was first asked in November 2023.

• When asked to what extent they trust the government to use AI to complete some of its tasks, 29% of adults strongly agreed or agreed, 27% neither agreed nor disagreed, and 43% strongly disagreed or disagreed.

• Men were more likely than women to say they trusted the government to use AI to complete some of its tasks (35% compared with 24%).

• When asked which public services, if any, they would trust to use AI to carry out some of their tasks; transport (35%), healthcare (32%) and education (31%) were the most commonly trusted, while social care and defence (both 19%) had the lowest reported levels of trust.

• Around half of adults said that they felt comfortable using AI in public transport for identifying fraud (51%), and analysing data to inform policy (51%); while lower levels of comfort were reported for managing air traffic and flight routes (15%) and for use in self-driving vehicles (15%).

In the latest period (6 to 31 August 2025), the cost of living (87%) and the NHS (82%) continued to be the most commonly reported important issues facing the UK (Figure 1). The proportion of adults identifying the NHS as an important issue has remained broadly stable since October 2022 (82%). In contrast, the proportion reporting the cost of living as an important issue has gradually declined over the same period, from 93% in October 2022.

The proportions of adults reporting immigration (65%) and crime (60%) as important issues have increased since October 2022, when these proportions were 43% and 49%, respectively.

Figure 1: The cost of living, the NHS, and the economy have been the most commonly reported important issues facing the UK since October 2022

Proportion of adults reporting each important issue, Great Britain, October 2022 to August 2025

Notes:

1. Respondents could select more than one option.
2. The length of each data collection period presented in this time series may be different.

Download the data

Public perceptions of AI benefits

In the latest period (6 to 31 August 2025), around 2 in 5 adults (41%) agreed that the use of artificial intelligence (AI) would benefit them (8% strongly agreed and 33% agreed). This proportion has remained broadly stable since the question was first asked in November 2023 (Figure 2). Around 1 in 5 adults (20%) disagreed or strongly disagreed, while around 2 in 5 (39%) neither agreed nor disagreed.

The proportion of adults selecting “neither agree nor disagree” has decreased from 47% in November 2023 to 39% in the latest period.

Figure 2: The proportion of adults who think that AI will benefit them has generally remained stable since we started asking in November 2023

Proportion of adults reporting the extent to which they agree that artificial intelligence (AI) will benefit them, Great Britain, November 2023 to August 2025

Notes:

1. The definition of AI presented to respondents is available in the dataset provided with this release.

Download the data

The following groups were more likely to agree or strongly agree that AI will benefit them:

• younger adults aged 16 to 29 years (49%) and 30 to 49 years (48%) compared with older adults (50 to 69 years (36%), 70 years and over (26%))

• men (46%) compared with women (36%)

Older adults were more likely to select "neither agree nor disagree" when asked whether AI would benefit them. Over half (53%) of those aged 70 years and over, and 46% of those aged 50 to 69 years selected this response, compared with 34% of those aged 30 to 49 years, and 27% of those aged 16 to 29 years.

Trust in the use of AI in government

In the latest period, we asked adults in Great Britain to what extent they agree or disagree with the statement: "I trust the government to use artificial intelligence (AI) to complete some of its tasks." There has been no significant change in responses since this question was last asked in November 2024:

• 29% of adults strongly agreed (3%) or agreed (27%)

• 27% neither agreed nor disagreed

• 43% strongly disagreed (17%) or disagreed (26%)

Men were more likely than women to say they trusted the government to use AI to complete some of its tasks (35% compared with 24%).

Adults aged 30 to 49 years were the most likely to report trust in the government's use of AI, with 38% agreeing or strongly agreeing with the statement. This was higher than all other age groups: 27% of those aged 16 to 29 years, and 25% of those aged 50 to 69 years, and 70 years and over.

When asked which public services, if any, they would trust to use AI to carry out some of their tasks, more than half of adults in Great Britain (54%) said they would trust at least one public service (Figure 4), a slight decrease from 58% in November 2024.

The most commonly reported services were:

• transport (35%)

• healthcare (32%)

• education (31%)

The public services where adults were the least likely to report trust in the use of AI to complete some tasks were social care (19%) and defence (19%).

Figure 4: Transport, healthcare, and education were the public services most trusted by adults to use AI for some tasks

Proportion of adults reporting public services they would trust to use artificial intelligence (AI) to complete some of their tasks, Great Britain, 6 to 31 August 2025

Notes:

1. Respondents could select more than one option.
2. The definition of AI presented to respondents is available in the dataset provided with this release.
3. “At least one public service" refers to adults who selected that they would trust one or more public services to use AI to carry out some of their tasks.

Download the data

Men (59%) were more likely than women (49%) to say they would trust the use of AI to complete tasks in at least one of the public services we asked about.

AI in public transport

We also asked adults specifically about their trust in the use of AI in public transport. When asked whether they agreed with the statement "I trust the use of artificial intelligence (AI) in public transport to complete some of its tasks":

• 39% of adults agreed or strongly agreed (7% strongly agreed and 33% agreed)

• 30% neither agreed nor disagreed

• 31% disagreed or strongly disagreed (18% disagreed and 13% strongly disagreed)

Men were more likely than women to say they trusted the use of AI in public transport (43% compared with 37%).

Comfort with the use of AI in public transport

We asked adults in Great Britain which uses of AI in public transport (listed in Figure 5) they would feel comfortable with. This question was also asked in our November release for comparison.

The top four most commonly reported uses in public transport remained the same when compared with November 2024. These were:

• to identify fraud (51% compared with 53%)

• to analyse data to inform policy (51% compared with 49%)

• measuring the environmental impact of transport projects (42% compared with 44%)

• controlling traffic lights and congestion (40% compared with 43%)

The lowest levels of comfort were reported for AI managing air traffic and flight routes (15%) and for use in self-driving vehicles (15%). Around a quarter (23%) of adults said that they would not feel comfortable with any of the listed uses (Figure 5).

Figure 5: Around half (51%) of adults feel comfortable with AI being used to identify fraud, and analysing data to inform policy

Proportion of adults reporting which public services they would feel comfortable using AI to complete some specific tasks, Great Britain, 6 to 31 August 2025

Notes:

1. Respondents could select more than one option.
2. The definition of AI presented to respondents is available in the dataset provided with this release.

Download the data

In the latest period, we also asked which additional uses of AI in transport services people felt comfortable with. Almost 7 in 10 (68%) said they would feel comfortable with AI being used to update train timetables in real time during disruptions.

Around a quarter supported AI issuing penalties and warnings for driving offences (24%) or reviewing a person's eligibility for a driving licence (24%). Around 1 in 5 adults (20%) said that they would not feel comfortable with any of the listed uses (Figure 6).

Figure 6: Around two-thirds (68%) of adults feel comfortable with AI being used to update train timetables in real time during disruptions

Proportion of adults reporting other public services they would feel comfortable using AI to complete specific tasks, Great Britain, 6 to 31 August 2025

Notes:

1. Respondents could select more than one option.
2. The definition of AI presented to respondents is available in the dataset provided with this release.

Download the data

Estimates presented in this bulletin are taken from the Public opinions and social trends, Great Britain: artificial intelligence dataset, published alongside this release.

Additional breakdowns of public attitudes towards artificial intelligence, including its use in government and public transport are available in a separate dataset based on pooled responses collected between July and August 2025. These breakdowns are presented by different population groups and provide further insight into views across demographic characteristics.

The analysis in this bulletin is based on adults aged 16 years and over in Great Britain.

In the latest period, 6 to 31 August 2025, we sampled 7,920 households. This sample was randomly selected from people who had previously completed the Labour Market Survey (LMS) or Opinions and Lifestyle Survey (OPN). The responding sample for the latest period contained 3,308 individuals, representing a 42% response rate.

Survey weights were applied to make estimates representative of the population, based on our population estimates. Estimates for some groups of the population may be subject to greater uncertainty because of smaller sample sizes for these groups (for example, younger adults).

From the 2 to 27 July 2025 data collection period onwards, the population totals used in weighting OPN estimates are based on 2022 mid-year estimates using updated population projections. For England and Wales, they are projected forward using scaling factors from 2021-based national population projections. For Scotland, they are projected forward using scaling factors from 2020-based national population projections.

For all estimates in the datasets, confidence intervals are provided. Where comparisons between estimates are made, associated confidence intervals should be used to assess the statistical significance of the differences. In some cases, additional statistical hypothesis testing was performed to identify differences between groups.

Further information on the survey design and quality can be found in our Opinions and Lifestyle Survey quality and methodology information (QMI).