Official Statistics

DSIT Public Engagement Survey 2025/2026

Published 16 July 2026

Key Findings

Digital inclusion and skills

In 2025/2026, the survey found that in the UK:

• the vast majority of adults (95%) reported that they had used the internet in the last three months. This very high share was seen across all age groups except those aged 75 to 84 and 85 and over, where usage fell to 78% and 46%, respectively  

• of those adults who used the internet, 98% had used the internet at home

• 76% felt that they could complete most digital tasks without difficulty. The proportion of adults that reported this decreases consistently with age, with only 13% of those aged 85 or over reporting that they could do this

• 56% of adults had improved their digital or online skills in the last 12 months. Self-teaching new digital skills was the most common single method, with 37% of adults adopting this style of learning 

• of adults wanting to improve their digital or online skills, saving time (45%), better organising their life (44%) and keeping up to date with digital skills (43%) were the three most common motivations

Internet connections and devices

In 2025/2026, the survey found that in the UK:

• broadband/home internet was the most common method of internet access at home, being used by 91% of adults 

• 31% of adults had connected to the internet at home using mobile data or hotspot; however, only 3% used this as their primary method of accessing the internet at home 

• of adults that connected to the internet at home, 39% felt their connection speed met their needs all of the time. Adults who considered they were living comfortably financially were most likely to have had their connection speed needs met all of the time (50%)

• of adults with a broadband connection, around half (49%) used superfast broadband (30 Mbps to 300 Mbps), 18% used ultrafast broadband (300 Mbps or higher) and 19% used standard broadband (below 30 Mbps). A further 14% did not know their broadband speed

• 96% of adults had at least one of the following devices available for personal use: a smartphone, laptop computer, tablet or desktop computer

• almost all adults (94%) had heard of 5G technology; however, only 60% reported that they used it

Government digital services

In 2025/2026, the survey found that in the UK:

• 78% of adults had accessed at least one government digital service in the last 12 months. The most common reason for having done so was to access services relating to driving and transport with around half (51%) using these services 

• of those that had accessed government digital services, 70% had done so via a website using a computer, laptop or tablet, while 60% had done so by using a website on a mobile phone. 39% had used some form of mobile app. Those aged 45 or over were more likely to access a government digital service website on a computer, laptop or tablet, and were less likely to have used a website on a mobile phone, compared to younger age groups

• 82% of adults using government digital services reported a positive experience. Around half (49%) of users of government digital services experienced no problems. The most common problem encountered was the overall process taking too long (16%), followed by adults being unable to remember a username or password (13%)

Attitudes towards science, technology and data

In 2025/2026, the survey found that in the UK:

• around three quarters (77%) of adults showed some level of interest in science, with 28% feeling connected to science and actively seeking out information. However, almost a quarter (23%) felt that science was not for them  

• those aged 16 to 54 tended to feel more connected with science and actively sought out further information compared to those that were aged 55 or over. Adults aged 85 or over were more likely to feel that science was not for them (54%). Those who identified as male were more likely to feel connected to science (37%) compared to those who identified as female (20%)

• adults were more comfortable with the public sector (75%) and government (72%) using data to uncover patterns and trends than businesses (54%) using data in this way. Generally, adults reported more similar levels of comfort when they thought about how the public sector (79%), government (77%) and businesses (77%) used the data they collected to make better decisions and deliver services.

• there was strong public support for scientists playing an active role in public policy, with the majority of adults agreeing that scientists should take an active role in informing public policy debates (73%)

Artificial intelligence (AI)

In 2025/2026, the survey found that in the UK:

• almost all adults (97%) reported they were at least aware of AI

• AI usage varies by type, with the most common uses being using AI that creates human-like text or speech in response to queries (used by 56% of adults in the last three months) and AI-powered digital assistants that can understand natural language (used by 54% of adults in the last three months)

• 59% of adults had used a form of generative AI in the last three months. Usage was most common in adults aged 16-44 and decreased with age after this point, down to 8% for those that were aged 85 or over

• use of generative AI varied geographically. The highest rates of usage were seen in London, where 69% had reported using this in the last three months. Usage was lowest in North East England (51%)

• of adults using generative AI in the online sample, 83% had used this in a personal setting and 54% for work purposes 

• the single largest reported perceived benefit of using AI was that it can allow easier access to information or advice (47%). However, 74% were concerned that the information produced by AI might be inaccurate

1. Introduction

The Department for Science, Innovation and Technology (DSIT) Public Engagement Survey is a nationally representative annual survey of adults (aged 16+) in the UK that tracks the latest trends in engagement with DSIT sectors. This release provides estimates for people who engaged with, or experienced, science and technology in their day-to-day lives, reported during the period of November 2025 to March 2026 (2025/2026). The format of the survey is push-to-web, with a paper version available for those who were unable to complete the survey online or would prefer an offline alternative.   

This is the first year of the survey and so results are presented for the period November 2025 to March 2026 only. Time series comparisons will be included once available in statistical releases for subsequent survey years.  

Official statistics for digital inclusion and skills, and broadband and mobile data usage were previously published by the Department for Culture, Media and Sport (DCMS) until 2024/25 as part of the Participation Survey. Due to differences in methodology and changes in question design, indicators for topics covered in both the 2024/25 Participation Survey and the DSIT Public Engagement Survey 2025/2026 should not be considered as being directly comparable. Statistics reporting the non-digital elements of the Participation Survey continue to be published by DCMS as part of the new Community and Engagement Survey.  

This report presents the headline estimates from 2025/2026, alongside demographic and geographic breakdowns. Full details of fieldwork dates can be found in the accompanying technical report. Further detailed estimates can be found in the accompanying data tables.

1.1 Code of Practice for Statistics 

The DSIT Public Engagement Survey results have been used to produce official statistics that adhere to the standards set out in the Code of Practice for Statistics.  

The survey is in its first year and we welcome feedback from users on the content and presentation of the survey results.  

The responsible statistician for this release is George Pickering. For enquiries on this release, please contact DSIT at publicengagementsurvey@dsit.gov.uk.

1.2 Methodology 

The 2025/2026 DSIT Public Engagement Survey was conducted via an online or paper questionnaire using Address Based Online Surveying (ABOS), this method is also sometimes referred to as “push-to-web” surveying.  Addresses were randomly sampled, with stratification by International Territorial Level 2 (ITL2) region, from the Royal Mail’s Postcode Address File (PAF).   

All sampled addresses were sent an initial invitation letter containing login details for up to four adults (16+) to complete the online survey. Two reminder letters were sent to all addresses where not every adult had responded, the first reminder was sent two weeks after the initial invitation and another four weeks after the initial invitation. A subset of these addresses also received a third reminder six weeks after the first letter.

As well as the online survey, respondents were given the option to complete a paper questionnaire on request. In addition, paper questionnaires were proactively provided to some addresses with the second reminder letter. These were targeted at sampled addresses in the most deprived quintile group and sampled addresses where it was expected that every resident would be aged 65 or over. Paper questionnaires were also enclosed for a random subset of addresses outside the two groups targeted at this stage.

Fieldwork for the 2025/2026 DSIT Public Engagement Survey was conducted between November 2025 and March 2026. The fieldwork period was divided into three batches, beginning on the following dates: 

• Batch 1: 18 November 2025 

• Batch 2: 06 January 2026 

• Batch 3: 03 February 2026 

Responses were accepted from all batches until the end of the fieldwork period on 31 March 2026. 

In total 83,694 addresses were sampled, from which 32,175 respondents completed the survey. 28,575 of these were via the online survey and 3,600 by completing a paper questionnaire. Following data quality checks, 1,477 respondents were removed, leaving 30,698 respondents in the final dataset. 

Fieldwork performance indicators are provided in table 1.1. Response rates are described in the following ways:

• household response rate - the percentage of households contacted as part of the survey from which at least one questionnaire was completed (25%)

• individual response rate - the estimated percentage of adults within sampled households that responded to the survey (21%). This is derived based on the Office for National Statistics’ Labour Force Survey estimate of the average number of adults per residential household

Both response rate indicators were adjusted to account for the estimated proportion of non-residential addresses within the PAF (8%). 

Table 1.1: Fieldwork performance indicators by batch

	Number of sampled addresses [footnote 1]	Complete responses (online and paper)	Number of households completed	Household response rate	Individual response rate [footnote 2]
Batch 1	26,656	9,312	5,899	24%	20%
Batch 2	33,315	13,099	8,344	27%	23%
Batch 3	23,723	8,287	5,309	24%	20%
Total	83,694	30,698	19,552	25%	21%

Full details of the methodology used to produce these statistics are available in the accompanying technical report.

Key caveats

Estimates in this report are rounded to the nearest whole number, with a few exceptions that are rounded to one decimal place. Estimates rounded to three decimal places are available in the corresponding data tables.

Some survey questions are multi-coded meaning that respondents could select any number of response options. In these cases, the sum of the percentages for responses will exceed 100%. 

The DSIT Public Engagement Survey issued both web and paper questionnaires. In most sections, the questionnaires have identical content, but in a few cases, questions with complex survey routing have been excluded from the paper questionnaire. This mostly affects more detailed questions on how AI has been used. Questions relating to whether AI had been used at all, awareness of AI, and perceptions of risks and benefits were asked in both versions of the questionnaire. Notes have been added where questions were only asked to online respondents, and where the statistics produced will only reflect the online population.

Confidence intervals

The upper and lower bounds displayed in the (single and multiple) bar charts in this report have been calculated using a 95% confidence interval. This means that, if the survey were repeated 100 times, 95 of the calculated confidence intervals would be expected to contain the true value for the group or geography. The smaller the sample size, the larger the confidence interval, as there is greater uncertainty about how well the sample represents the wider population. As a result, it is more difficult to draw inferences from these results.

2. Digital inclusion and skills

The survey asked adults questions relating to digital skills and inclusion, such as how frequently they accessed the internet or if they had improved their digital or online skills in the last 12 months. The full data tables can be found for this chapter in the accompanying ‘DSIT Public Engagement Survey 2025/2026: digital inclusion and skills tables’.

Throughout this section the term “internet user” is used, this is defined as someone who has reported that they had used the internet in the last three months in any location.

2.1 Digital inclusion

Figure 2.1 shows the proportion of adults that had used the internet in the last three months, by age group. The UK total is included for comparison.

Figure 2.1: Usage of the internet in the last three months, by age, UK, 2025/2026 (base: all adults)

Notes: data from figure 2.1 can be found in table A3 of the accompanying ‘DSIT Public Engagement Survey 2025/2026: digital inclusion and skills tables’.

Overall, 95% of adults in the UK had used the internet in the last three months. When broken down by age, similar figures were found across all age groups aged under 65. For 75 to 84 year olds, the proportion of internet users fell to 78%. The 85 and over age group had the lowest levels of internet usage, with 46% having used the internet in the last three months.  

Adults reported how often, if ever, they used the internet. 86% of adults used the internet every day, followed by most days (7%), and once or twice a week (2%). Around 4% of adults had never used the internet, with a further 1% having used but not in the last three months. Less than 1% of adults used the internet around once a month, less than once a month and in the past but not in the past three months.

Figure 2.2 shows adults’ self-reported satisfaction with the extent of their online activity, broken down by age (16 to 64 and 65 or over). The UK total is included for comparison.

Figure 2.2: Satisfaction with extent of online activity, by age, UK, 2025/2026 (base: all adults)

Notes: data from figure 2.2 can be found in table A5 of the accompanying ‘DSIT Public Engagement Survey 2025/2026: digital inclusion and skills tables’.

Overall, 92% of adults went online as much as they wanted, 2% went online but wanted to go online more, 1% did not go online but wanted to, and 5% did not go online and did not want to go online. For those aged 16 to 64, around 97% went online as much as they wanted. For those aged 65 or over this proportion fell to 79%, while 17% of adults aged 65 or over reported they did not go online and did not want to, with 2% not going online but wanting to.

Figure 2.3 shows where internet users in the UK accessed the internet by any means (Wi-Fi, mobile data, etc). It should be noted that these are percentages of all internet users and therefore variation will be strongly influenced by where people spend their time (e.g. workplaces, educational establishments etc).

Figure 2.3: Locations of internet access, UK, 2025/2026 (base: internet users)

Notes: data from figure 2.3 can be found in table A7 of the accompanying ‘DSIT Public Engagement Survey 2025/2026: digital inclusion and skills tables’.

Of internet users, 98% accessed the internet at home and similar findings are seen across all demographic groups. This is followed by on the move (59%), at work (55%) and in cafes/restaurants/pubs/bars (54%). When broken down by different age groups, the proportion of adults who accessed the internet in various locations is similar for groups aged between 16 and 54. 16 to 24 year olds accessed the internet in libraries, schools, colleges or universities much more than all other age groups. From age 45 onwards, the proportions of adults who accessed the internet at almost all locations away from home fell as age increased.

Adults were also asked about locations away from home at which they connected to the internet using Wi-Fi specifically. Figure 2.4 shows the locations where adults using the internet away from home had connected using Wi-Fi.

Figure 2.4: Locations connected to the internet away from home using Wi-Fi, UK, 2025/2026 (base: internet users that accessed the internet away from home)

Notes: data from figure 2.4 can be found in table A9 of the accompanying ‘DSIT Public Engagement Survey 2025/2026: digital inclusion and skills tables’.

Overall, those using the internet away from home accessed the internet the most using Wi-Fi at work (55%). This was followed by cafes, restaurants, pubs and bars (48%) and in someone else’s home (47%).

2.2 Digital skills

Figure 2.5 shows the proportion of internet users in the UK that had undertaken a range of digital tasks using the internet in the last three months without help from someone else.

Figure 2.5: Digital tasks undertaken in the last three months without help, UK, 2025/2026 (base: internet users)

Notes: data from figure 2.5 can be found in table A11 of the accompanying ‘DSIT Public Engagement Survey 2025/2026: digital inclusion and skills tables’.

The proportion of internet users who had looked something up using a search engine was 97%, making it the most common digital task. This was followed by sending or receiving emails (96%) and sending or receiving messages using apps (e.g. on WhatsApp, Messenger), where again 96% had done this task.  

A high share of internet users had used it for online banking (92%) and buying products or services online (92%). 

Booking travel or holidays online was the least common digital task. 66% of internet users had done this.

Figure 2.6 shows how all adults in the UK self-reported their current digital skills, considering the extent to which they were able to complete digital tasks without encountering difficulties or requiring assistance. The UK total is included for comparison.

Figure 2.6: Self-reported current digital skills, by age, UK, 2025/2026 (base: all adults)

Notes: data from figure 2.6 can be found in table A13 of the accompanying ‘DSIT Public Engagement Survey 2025/2026: digital inclusion and skills tables’.

Overall, 76% of adults reported that they could complete most digital tasks without difficulty, 15% that they could complete some digital tasks, but struggled with harder ones, 3% felt that they could only do basic tasks and needed help often and 2% struggled with most digital tasks. A high proportion of adults in each age group up to age 44 reported they could complete most digital tasks without difficulty (88% to 94%). For adults aged 45 and over, this proportion decreased as age increased, with adults aged 85 or over least likely to be able to complete most digital tasks without difficulty (13%). Within the 85 or over age group, 10% felt that they were only able to complete basic tasks and often needed help and 15% struggled with most digital tasks. 

Figure 2.7 shows adults’ self-reported current digital skills by disability status in the UK. Disability in the survey was defined if the respondent answered both “yes” to the question “Do you have any physical or mental health conditions or illnesses lasting or expected to last for 12 months or more?” and then “yes, a lot” or “yes, a little” to the follow-up question “Do any of these conditions or illnesses reduce your ability to carry out day-to-day activities?”.

Figure 2.7: Self-reported current digital skills, by disability status, UK, 2025/2026 (base: all adults)

Notes: data from figure 2.7 can be found in table A13 of the accompanying ‘DSIT Public Engagement Survey 2025/2026: digital inclusion and skills tables’.

Adults with a disability reported that they were less likely to be able to complete most digital tasks without difficulty (59%) compared to adults without a disability (82%).

Figure 2.8 shows the reasons why adults who either do not use the internet or would like to go online more did not use the internet.

Figure 2.8: Reasons for not using the internet, UK, 2025/2026 (base: adults who either do not use the internet or would like to go online more)

Notes: data from figure 2.8 can be found in table A15 of the accompanying ‘DSIT Public Engagement Survey 2025/2026: digital inclusion and skills tables’. In the online questionnaire this question was asked to adults that either did not use the internet or did but wanted to use it more. In the paper questionnaire this question was asked to adults that did not use the internet only.

Overall, the most common reason reported by adults for not using the internet was that they did not have the skills or knowledge (39%). Other common reasons included preferring to do things in person or over the phone (35%), being able to ask someone else to do online tasks for them (31%) and being worried about scams or harmful content online (30%). Some of the least common reasons included adults not having anyone to help them (8%), having a disability that made it hard to use the internet (10%) and due to poor broadband or mobile coverage in their area (5%). Around 13% did not use the internet because it was too expensive.

2.3 Digital and online skills training

This section focuses on digital and online skills training that adults had either done or were interested in doing. Survey responses were self-reported, therefore digital and online skills training will include a wide range of activities at varying degrees of difficulty. It also covers the perceived benefits associated with digital or online skills training.

Figure 2.9 shows the proportion of adults that had improved their digital or online skills in the last 12 months using different methods of upskilling.

Figure 2.9: Methods used to improve digital or online skills in the last 12 months, UK, 2025/2026 (base: all adults)

Notes: data from figure 2.9 can be found in table A17 of the accompanying ‘DSIT Public Engagement Survey 2025/2026: digital inclusion and skills tables’.

The most common method of improving digital or online skills was by self-taught methods (37%). Learning new digital skills from family and friends was the second most common way adults improved their digital skills (23%). Taking part in structured digital skills training outside of work which was free (4%) or paid for (2%) had the two lowest proportions of uptake.

Figure 2.10 shows the proportion of adults who had improved their digital or online skills using at least one method in the last 12 months, broken down by age group. The UK total is included for comparison.

Figure 2.10: Adults who had improved their digital or online skills in the last 12 months, by age, UK, 2025/2026 (base: all adults)

Notes: data from figure 2.10 can be found in table A17 of the accompanying ‘DSIT Public Engagement Survey 2025/2026: digital inclusion and skills tables’.

Overall, 56% of adults in the UK had improved their digital or online skills in the last 12 months. At least half of the adults aged 16 to 64 had improved their digital or online skills in the last 12 months. Of these adults, 16 to 19 year olds were the most likely to have improved their digital or online skills (66%). Adults aged 85 or over were the least likely to have improved their digital or online skills in the last 12 months (24%).

Figure 2.11 is a map of UK regions and countries at International Territorial Level 1 (ITL1), displaying the proportion of adults who improved their digital or online skills in the last 12 months. It includes participation in any form of digital upskilling or training.

Figure 2.11: Adults who had improved their digital or online skills in the last 12 months, by ITL1 area, UK, 2025/2026 (base: all adults)

Notes: data from figure 2.11 can be found in table A18 of the accompanying ‘DSIT Public Engagement Survey 2025/2026: digital inclusion and skills tables’.

Adults in London (60%), Northern Ireland (58%) and the East of England (57%) were the most likely to have improved their digital or online skills in the last 12 months. Conversely, adults in Yorkshire and the Humber (53%), Scotland (52%) and the North East of England (52%) were the least likely to have improved their digital or online skills in the last 12 months.

Adults were also asked about their future intentions for improving their digital or online skills. Figure 2.12 shows the interest that adults in the UK had in participating in different ways of improving their digital or online skills. It does not reflect digital or online skills training that they had already completed.

Figure 2.12: Interest in different methods of digital or online skills training, UK, 2025/2026 (base: all adults)

Notes: data from figure 2.12 can be found in table A19 of the accompanying ‘DSIT Public Engagement Survey 2025/2026: digital inclusion and skills tables’.

Overall, in future, adults were most interested in teaching themselves (47%), learning through friends and family (26%) and through work (23%). There was less interest in learning through organised group learning sessions/evening classes (8%), local support centres (8%) and one-to-one lessons (7%). Nearly two‑thirds of adults (62%) were interested in at least one form of digital upskilling.

Figure 2.13 is a map of UK ITL1 regions and countries, displaying the proportion of adults who were interested in improving their digital or online skills. It includes interest in any form of digital upskilling.

Figure 2.13: Interest in improving digital or online skills, by ITL1 area, UK, 2025/2026 (base: all adults)

Notes: data from figure 2.13 can be found in table A20 of the accompanying ‘DSIT Public Engagement Survey 2025/2026: digital inclusion and skills tables’.

Adults in Northern Ireland (69%), London (67%) and the East of England (63%) were the most likely to be interested in improving their digital or online skills. Adults in the North East of England (54%), the West Midlands (60%) and Scotland (60%) were the least likely to be interested in improving their digital and online skills.

Figure 2.14 shows what adults who were interested in improving their digital or online skills thought the benefits of digital upskilling were.

Figure 2.14: Perceived benefits of improving digital or online skills, UK, 2025/2026 (base: adults interested in improving digital or online skills)

Notes: data from figure 2.14 can be found in table A21 of the accompanying ‘DSIT Public Engagement Survey 2025/2026: digital inclusion and skills tables’.

Adults were most likely to think that saving time (45%), organising their life better (44%), keeping up to date with digital skills (43%) and improving skills for their job (41%) were benefits of improving their digital or online skills. Feeling more like part of a community (10%) and managing and improving health (19%) were the least likely perceived benefits of digital upskilling.

3. Internet connections and devices

The survey asked adults questions relating to how they connected to the internet at home and which devices they had available for personal use. The full data tables can be found for this chapter in the accompanying ‘DSIT Public Engagement Survey 2025/2026: internet connections and devices tables’.

3.1 Internet connections

Figure 3.1 presents the single main method used to connect to the internet at home, reported by adults that had internet access at home.

Figure 3.1: Main method of connecting to the internet at home, UK, 2025/2026 (base: adults who used the internet at home)

Notes: data from figure 3.1 can be found in table B3 of the accompanying ‘DSIT Public Engagement Survey 2025/2026: internet connections and devices tables’.

A strong majority of adults that used the internet at home (92%) said broadband was their main method of connecting to the internet (91% of all adults had connected to the internet at home this way). A further 4% connected via a Wi-Fi dongle or router that provided home internet via a mobile network. Although 31% of adults overall had connected to the internet at home using mobile data or hotspot, only 3% of those using the internet at home reported this as their primary method of connection.

Figure 3.2 shows the ways that adults who use the internet at home pay for their home internet connection, broken down by age. The UK total is included for comparison.

Figure 3.2: How internet or broadband at home is paid for, by age, UK, 2025/2026 (base: adults who used the internet at home)

Notes: data from figure 3.2 can be found in table B5 of the accompanying ‘DSIT Public Engagement Survey 2025/2026: internet connections and devices tables’.

Overall, 54% of adults that used the internet at home paid for internet or broadband by itself. However, for adults aged 55 and over, packages/bundles were, on average, the most used way of paying for their broadband or internet. For adults aged under 55, paying for internet or broadband only was most common.

Figure 3.3 shows the different types of broadband services paid for, by adults that had a broadband connection at home. These were defined in terms of download speed, measured in megabits per second (Mbps): 

• standard broadband, below 30 Mbps 

• superfast broadband, 30 Mbps to 300 Mbps 

• ultrafast broadband, 300 Mbps or higher

Figure 3.3: Household broadband speed, UK, 2025/2026 (base: adults with a broadband connection)

Notes: data from figure 3.3 can be found in table B7 of the accompanying ‘DSIT Public Engagement Survey 2025/2026: internet connections and devices tables’.

Of the adults with a broadband connection, 49% used superfast broadband, 19% used standard broadband and 18% used ultrafast broadband.

Figure 3.4 shows the speed of broadband services used by adults with a broadband connection, broken down by household income.

Figure 3.4: Broadband speed, by household income, UK, 2025/2026 (base: adults with a broadband connection)

Notes: data from figure 3.4 can be found in table B7 of the accompanying ‘DSIT Public Engagement Survey 2025/2026: internet connections and devices tables’.

Superfast broadband (30 Mbps to 300 Mbps) makes up the largest share in every income band. Households with higher incomes are more likely to have ultrafast broadband (speed 300 Mbps or higher). The proportion using ultrafast broadband rises steadily as income increases and is most common among households earning £150,000 or more per year (33%). By contrast, adults in the lowest household income bands are most likely to rely on standard broadband (below 30 Mbps).

Figure 3.5 shows the extent to which adults perceived their home internet connection speed to be sufficient in meeting their needs, broken down by financial hardship. The UK total is included for comparison. Financial hardship in the survey was defined by the response given to the question “How well would you say you are managing financially these days?”. This question was asked to adults with any form of internet connection at home, not just broadband.

Figure 3.5: Extent to which needs are met by home internet connection speed, by financial hardship, UK, 2025/2026 (base: adults who use the internet at home)

Notes: data from figure 3.5 can be found in table B9 of the accompanying ‘DSIT Public Engagement Survey 2025/2026: internet connections and devices tables’.

Overall, 39% of adults who used the internet at home felt their connection speed met their needs all of the time, 49% felt they were met most of the time, 6% had their needs met about half of the time, 2% had their needs met only a little of the time and 1% none of the time. Adults who were living comfortably financially were the most likely to have their connection speed meet their needs all of the time (50%). As financial hardship increased this proportion decreased, with adults finding it very difficult financially being the least likely to have their connection speed meet their needs all of the time (28%).

3.2 Devices and mobile data

Figure 3.6 shows the prevalence of different devices that adults had available for personal use. Devices in scope of this question included smartphones, tablets, laptop computers and desktop personal computers (PC).

Figure 3.6: Devices available for personal use, UK, 2025/2026 (base: all adults)

Notes: data from figure 3.6 can be found in table B13 of the accompanying ‘DSIT Public Engagement Survey 2025/2026: internet connections and devices tables’.

Overall, 93% of adults had access to a smartphone for personal use, 67% had a laptop computer, 52% had a tablet (e.g. iPad) and 24% had a desktop PC. 96% of adults had at least one of these devices available for personal use.

Figure 3.7 shows adults’ access to these devices for personal use, by age group (16 to 64 and 65 or over).

Figure 3.7: Devices available for personal use, by age, UK, 2025/2026 (base: all adults)

Notes: data from figure 3.7 can be found in table B13 of the accompanying ‘DSIT Public Engagement Survey 2025/2026: internet connections and devices tables’.

Adults aged 16 to 64 were more likely to have access to a smartphone (98% compared to 80%) and laptop (75% compared to 44%) than those aged 65 or over. Around half of both age groups had access to a tablet. 16 to 64 year olds were about as likely to have a desktop PC available for personal use (24%) as those aged 65 or over (23%). 12% of adults aged 65 or over had access to none of these devices, compared to just 1% for adults aged 16 to 64.

Adults with access to a smartphone were then asked a series of questions relating to how they paid for this and their use of mobile data. Figure 3.8 shows the monthly price paid for a smartphone by adults in the UK, excluding those who had a mobile package that was bundled with internet and broadband. Responses to this question are reflective of both those paying for mobile data only, and those paying for a package including a phone.

Figure 3.8: Monthly price paid for smartphone, UK, 2025/2026 (base: adults who pay for a mobile package which is not bundled with internet/broadband)

Notes: data from figure 3.8 can be found in table B21 of the accompanying ‘DSIT Public Engagement Survey 2025/2026: internet connections and devices tables’.

Adults were most likely to pay between £6 and £10 per month (25%), followed by between £11 and £15 (19%) and between £21 and £30 (14%). Adults were least likely to pay more than £60 per month (2%).

Figure 3.9 shows adults’ monthly data allowance for their most used smartphone.

Figure 3.9: Monthly data allowance for most used mobile phone, UK, 2025/2026 (base: adults with a smartphone)

Notes: data from figure 3.9 can be found in table B23 of the accompanying ‘DSIT Public Engagement Survey 2025/2026: internet connections and devices tables’.

Most adults had a capped monthly data allowance but 23% of adults had unlimited data. The most common capped monthly data allowance was 10 to less than 30 gigabytes (GB) (18%) followed by 30 GB to 100 GB (14%). The least common responses were having no data (1%) and a monthly data allowance of less than 1 GB (1%).

Figure 3.10 shows how frequently adults with a smartphone ran out of mobile data, broken down by financial hardship. The UK total is included for comparison. This question did not include smartphone users that reported that they paid for an unlimited data package.

Figure 3.10: Frequency of running out of mobile data, by financial hardship, UK, 2025/2026 (base: adults with a smartphone who do not have unlimited data)

Notes: data from figure 3.10 can be found in table B25 of the accompanying ‘DSIT Public Engagement Survey 2025/2026: internet connections and devices tables’.

Overall, most adults without unlimited data packages never run out of mobile data (69%). Around 19% of adults occasionally run out of mobile data, with 9% running out at least once every three months. As financial hardship increased, the proportion of adults who ran out of mobile data every month increased. Adults who were finding it very difficult financially were most likely to run out of mobile data every month (8%).

All survey participants were asked about their awareness of fifth generation (5G) mobile technology. Figure 3.11 shows the level of awareness by UK adults, broken down by age and including the UK total for comparison.

Figure 3.11: Awareness of 5G mobile technology, by age, UK, 2025/2026 (base: all adults)

Notes: data from figure 3.11 can be found in table B27 of the accompanying ‘DSIT Public Engagement Survey 2025/2026: internet connections and devices tables’.

Overall, almost all adults (94%) had at least heard of 5G technology, 60% already used it, 13% knew what it was and were interested in getting it in the near future, 10% knew what it was, but were not interested in getting it in the near future, and 12% had heard of it but were not sure what it was. Only 6% had not heard of 5G.  

The proportion of adults who had heard of 5G before increased with age, up until the 25 to 34 age group (99%). It then decreased with age, with those aged 85 or over having a significantly larger proportion who had not heard of 5G before (46%) compared to other age groups. 

The proportion of those who were already using it increased with age, from 65% (16 to 19 year olds) to 80% (25 to 34 year olds). It then steadily decreased with age, with those aged 85 or over having the lowest proportion of any age group (7%).

4. Government digital services

The survey asked adults questions relating to government digital services. Government digital services include the GOV.UK website or app, as well as services accessed via digital channels, such as apply for a passport and Universal Credit. It also includes services accessed through other UK Government apps such as the HMRC app, GOV.UK ID Check and GOV.UK One Login but excludes other public sector apps e.g. NHS.

The full data tables can be found for this chapter in the accompanying ‘DSIT Public Engagement Survey 2025/2026: government digital services tables’.

4.1 Use of government digital services

Figure 4.1 shows adults’ usage of different categories of government digital services in the last 12 months.

Figure 4.1: Usage of government digital services in the last 12 months, UK, 2025/2026 (base: all adults)

Notes: data from figure 4.1 can be found in table C1 of the accompanying ‘DSIT Public Engagement Survey 2025/2026: government digital services tables’.

Driving and transport was the most commonly used category of government digital service (51%), followed by services for passports, travel and living abroad (26%), money and tax (25%), and housing and local services (19%).

A survey respondent was defined as being a user of government digital services if they reported that they had used at least one government digital service in the last 12 months. Figure 4.2 shows the proportion of adults that had used at least one government digital service in the last 12 months, broken down by age and including the UK total for comparison.

Figure 4.2: Usage of at least one government digital service in the last 12 months, by age, UK, 2025/2026 (base: all adults)

Notes: data from figure 4.2 can be found in table C3 of the accompanying ‘DSIT Public Engagement Survey 2025/2026: government digital services tables’.

Overall, 78% of adults had accessed at least one government digital service in the last 12 months.  Adults aged 85 or over (32%), 75 to 84 (56%) and 16 to 19 (66%) were least likely to have used at least one government digital service in the last 12 months. The remaining age groups all had similar levels of engagement with at least one government digital service, falling between 74% (65 to 74 year olds) to 88% (34 to 45 year olds).

Users of government digital services were then asked about their experiences of accessing and using these services. Figure 4.3 shows the methods used by adults to access government digital services in the last 12 months.

Figure 4.3: Methods of accessing government digital services in the last 12 months, UK, 2025/2026 (base: users of government digital services)

Notes: data from figure 4.3 can be found in table C5 of the accompanying ‘DSIT Public Engagement Survey 2025/2026: government digital services tables’.

Adults accessed government digital services via a website using a computer, laptop or tablet the most (70%). The second most common method was via a website using a phone (60%), followed by other UK Government app (20%) and the GOV.UK app (19%).

Figure 4.4 shows the method used to access government digital services in the last 12 months by age group.

Figure 4.4: Methods of accessing government digital services in the last 12 months, by age, UK, 2025/2026 (base: users of government digital services)

Notes: data from figure 4.4 can be found in table C5 of the accompanying ‘DSIT Public Engagement Survey 2025/2026: government digital services tables’.

For 16 to 44 year olds, the most common method of accessing government digital services was via a website using their phone (ranging from 67% to 75% depending on age group). Accessing a government digital service via a website using their phone became less common for those aged 45 and over, with adults aged 85 or over using this method the least (14%). For adults aged 45 and over, the most common method of accessing government digital services was via a website using a computer, laptop or tablet (66% to 76% depending on age group). Usage of the GOV.UK app increased as age increased from 16 to 19 year olds (15%) to 55 to 64 year olds (21%). This then began to decrease as age increased, reaching the lowest levels for adults aged 85 or over (8%).

4.2 Experiences of using government digital services

Users of government digital services were asked to rate their overall experience of using these services. When doing this they were asked to think about their overall experience of using the service itself, regardless of the outcome of the service.

Figure 4.5 shows adults’ overall experience when using government digital services in the last 12 months by age group. The UK total is included for comparison.

Figure 4.5: Overall experience of using government digital services in the last 12 months, by age, UK, 2025/2026 (base: users of government digital services)

Notes: data from figure 4.5 can be found in table C7 of the accompanying ‘DSIT Public Engagement Survey 2025/2026: government digital services tables’.

Adults in the UK largely had positive experiences of using government digital services, with 82% reporting either very good or fairly good experiences. The trend by age group shows that overall positive experiences increased as age group increased from those aged 16 to 19 (78%) to those aged 55 to 64 (86%). Positive experiences began to fall for those aged 65 and over, with those aged 85 or over having the lowest proportion of positive experiences (61%).

Figure 4.6 shows adults’ overall experience using government digital services in the last 12 months by disability status. Disability in the survey was defined if the respondent answered both “yes” to the question “Do you have any physical or mental health conditions or illnesses lasting or expected to last for 12 months or more?” and then “yes, a lot” or “yes, a little” to the follow-up question “Do any of these conditions or illnesses reduce your ability to carry out day-to-day activities?”.

Figure 4.6: Overall experience of using government digital services in the last 12 months, by disability status, UK, 2025/2026 (base: users of government digital services)

Notes: data from figure 4.6 can be found in table C7 of the accompanying ‘DSIT Public Engagement Survey 2025/2026: government digital services tables’.

Adults without a disability were more likely to have had a very good experience of using government digital services (37%), compared to adults with a disability (30%). Fairly good, fairly poor and very poor experiences were similarly reported between the two groups.

Figure 4.7 is a map of UK regions and countries at ITL1, displaying the proportion of users of government digital services in the last 12 months that reported a very good experience of using services.

Figure 4.7: Adults who reported their experience of using government digital services to be ‘very good’ in the last 12 months, by ITL1 area, UK, 2025/2026 (base: users of government digital services)

Notes: data from figure 4.7 can be found in table C8 of the accompanying ‘DSIT Public Engagement Survey 2025/2026: government digital services tables’.

The proportion of adults UK-wide who had a very good experience using government digital services was 35%. The area with the largest proportion of very good experiences was Northern Ireland (39%), followed by Wales (38%) and the West Midlands (37%). London had the lowest proportion of very good experiences (32%).

Users of government digital services were also asked about any problems experienced when using services. Figure 4.8 shows problems experienced by government digital services users in the last 12 months.

Figure 4.8: Problems experienced while using government digital services in the last 12 months, UK, 2025/2026 (base: users of government digital services)

Notes: data from figure 4.8 can be found in table C11 of the accompanying ‘DSIT Public Engagement Survey 2025/2026: government digital services tables’.

Around half (49%) of users of government digital services experienced no problems. The most common problem encountered was the overall process taking too long (16%), followed by adults being unable to remember a username or password (13%), the process being too difficult to complete (11%), experiencing technical difficulties (10%), not understanding the instructions or what to do (10%) and not having all the information that was needed (9%). The two least common problems were needing help, but none being available at the time (8%) and not knowing how to find help (7%).

5. Attitudes towards science, technology and data

The survey asked adults about their attitudes towards science, technology and data. This included questions about how comfortable they were with different organisations using their data and their relationship with science.

The full data tables can be found for this chapter in the accompanying ‘DSIT Public Engagement Survey 2025/2026: attitudes towards science, technology and data tables’.

Figure 5.1 shows adults’ self-reported level of interest in science by age, with the UK total included for comparison. This question was asked on a scale from feeling “connected with science” to feeling that “science is not for me”, with a middle option of being interested in science but not making a special effort to keep informed.

Figure 5.1: Level of interest in science, by age, UK, 2025/2026 (base: all adults)

Notes: data from figure 5.1 can be found in table D1 of the accompanying ‘DSIT Public Engagement Survey 2025/2026: attitudes towards science, technology and data tables’.

Overall, for the UK, 28% felt connected with science and actively sought out additional information relating to this, 49% had an interest in science but didn’t make a special effort to keep informed and 23% felt that science was not for them. Those aged 16 to 64 tended to feel more connected with science and actively sought out further information compared to those aged 65 or over. Adults aged 85 or over were most likely to feel that science was not for them (54%). 

Those with a degree-level or above qualification were more likely to feel connected to science (40%), compared to those with another type of qualification (18%) and those with no qualifications (7%).   

Those who identified as male were more likely to feel connected to science (37%) compared to those who identified as female (20%).

Figure 5.2 shows adults’ attitudes towards whether data collection and analysis are good for society. This was in response to the question “Overall, how much do you agree or disagree that collecting and analysing data is good for society?”.

Figure 5.2: Responses to “Overall, how much do you agree or disagree that collecting and analysing data is good for society?”, UK, 2025/2026 (base: all adults)

Notes: data from figure 5.2 can be found in table D3 of the accompanying ‘DSIT Public Engagement Survey 2025/2026: attitudes towards science, technology and data tables’.

Overall, 64% agreed that collecting and analysing data is good for society (including 26% who strongly agreed), with 22% neither agreeing nor disagreeing. 8% disagreed that collecting and analysing data is good for society (including 3% who strongly disagreed).

Figure 5.3 shows adults’ level of comfort with different types of organisations (businesses, government, public sector organisations) using data to make better decisions, and using data to uncover patterns and trends.

Figure 5.3: Attitudes towards data collection and usage by businesses, government, and public sector organisations, UK, 2025/2026 (base: all adults)

Notes: data from figure 5.3 can be found in tables D5, D7 and D9 of the accompanying ‘DSIT Public Engagement Survey 2025/2026: attitudes towards science, technology and data tables’.

In general, across all three types of organisations adults were more comfortable with data being used to make better decisions than it being used to uncover patterns and trends.  

Adults were more comfortable with the public sector (75%) and government (72%) using data to uncover patterns and trends than businesses (54%) using data in this way.  

Generally, adults reported similar levels of comfort when they thought about how the public sector (79%), government (77%) and businesses (77%) used the data they collected to make better decisions and deliver services.

Figure 5.4 shows how much adults trusted different types of organisations to keep people’s personal data safe.

Figure 5.4: Trust in organisations to keep personal data safe, UK, 2025/2026 (base: all adults)

Notes: data from figure 5.4 can be found in table D11 of the accompanying ‘DSIT Public Engagement Survey 2025/2026: attitudes towards science, technology and data tables’.

Overall, the organisations with the highest levels of trust amongst adults in the UK to keep personal data safe were banks and other financial institutions (26% had a great deal of trust, 49% a fair amount of trust) and the NHS (24% had a great deal of trust, 49% had a fair amount of trust). Government departments had slightly less public trust (12% had a great deal of trust, 48% had a fair amount of trust) than these organisations to keep personal data safe. The organisations with the lowest levels of trust from adults in the UK to keep personal data safe were social media companies (3% had a great deal of trust, 15% had a fair amount of trust), high street retailers and supermarkets (4% had a great deal of trust, 23% had a fair amount of trust), and big technology companies (6% had a great deal of trust, 29% had a fair amount of trust).

Figure 5.5 shows how much adults trusted different sources to provide accurate information about science and technology.

Figure 5.5: Trust in sources to provide accurate information about science and technology, UK, 2025/2026 (base: all adults)

Notes: data from figure 5.5 can be found in table D13 of the accompanying ‘DSIT Public Engagement Survey 2025/2026: attitudes towards science, technology and data tables’.

All sources had over 50% of adults trusting them at least a fair amount to provide accurate information about science and technology, except for UK Government ministers or members of parliament (MPs) (7% had a great deal of trust, 31% had a fair amount of trust). Health charities such as Cancer Research (21% had a great deal of trust, 54% had a fair amount of trust) and scientists working at universities (28% had a great deal of trust, 47% a fair amount) were the most trusted to provide accurate information about science and technology.

Adults were asked how much they agreed or disagreed with scientists taking an active role in informing public policy debates and working with politicians to ensure that policy making is based on scientific findings. Figure 5.6 shows adults’ attitudes towards scientists’ role in public policy for science and technology.

Figure 5.6: Attitudes towards scientists’ role in public policy for science and technology, UK, 2025/2026 (base: all adults)

Notes: data from figure 5.6 can be found in table D15 of the accompanying ‘DSIT Public Engagement Survey 2025/2026: attitudes towards science, technology and data tables’.

The majority of adults agreed that scientists should work with politicians to ensure that policymaking is based on scientific findings (40% strongly agreed, 37% tended to agree). Similarly, the majority of adults agreed that scientists should take an active role in informing public policy debates (34% strongly agreed, 40% tended to agree).

Figure 5.7 shows adults’ preferred methods for hearing about developments in science and health.

Figure 5.7: Preferred methods for hearing about developments in science and health, UK, 2025/2026 (base: all adults)

Notes: data from figure 5.7 can be found in table D17 of the accompanying ‘DSIT Public Engagement Survey 2025/2026: attitudes towards science, technology and data tables’. Up to a maximum of three response options could be selected for this question.

Overall, TV or radio news was the most preferred way of hearing about developments in science and health (59%). Newspapers or magazine articles (whether printed or online) (36%), social media posts (36%), TV, radio and streaming documentaries (33%) and a GP or health professional (32%) all had similar proportions of adults who preferred these methods. The methods with the lowest proportion of adults who preferred them were hearing from friends, family or colleagues (16%), podcasts (15%), not wanting to find out about science and health from anybody (4%) and hearing about developments in another way (2%).

Figure 5.8 shows the top three most commonly preferred methods to hear about developments in science and health for the UK overall (see figure 5.7), broken down by age group.

Figure 5.8: Selected preferred methods for hearing about developments in science and health, by age, UK, 2025/2026 (base: all adults)

Notes: data from figure 5.8 can be found in table D17 of the accompanying ‘DSIT Public Engagement Survey 2025/2026: attitudes towards science, technology and data tables’.

Younger age groups were more likely to prefer social media posts to hear about developments in science and health, whilst older age groups were more likely to prefer TV/radio news (such as the BBC) and newspaper or magazine articles. Adults were more likely to prefer TV/radio news compared with social media at age 35 or over, and this remained the most preferred way of hearing about developments in science and health for those age groups. Those aged 45 or over were more likely to prefer newspaper and magazine articles over social media posts.  

Some common methods for age groups are not displayed in figure 5.8. Adults aged 16 to 19 often preferred to hear about developments in science and health from friends, family or colleagues (25%). Hearing about developments in science and health via TV/radio/streaming documentaries was preferred for adults aged 45 to 54 (37%), 55 to 64 (41%) and 65 to 74 (37%). Adults aged 65 to 74 (46%), 75 to 84 (51%) and 85 or over (38%) were more likely to prefer hearing about developments in science and health from GP/health professionals.

6. Artificial intelligence (AI)

The survey asked adults questions about artificial intelligence (AI). This included questions about their awareness and usage of AI, as well as the perceived risks and benefits of the technology. The full data tables can be found for this chapter in the accompanying ‘DSIT Public Engagement Survey 2025/2026: artificial intelligence tables’.

6.1 AI adoption

Figure 6.1 shows the extent of awareness of AI by different age groups. Any form of AI was in scope for this question. The UK total is included for comparison.

Figure 6.1: Awareness of artificial intelligence (AI), by age, UK, 2025/2026 (base: all adults)

Notes: data from figure 6.1 can be found in table E1 of the accompanying ‘DSIT Public Engagement Survey 2025/2026: artificial intelligence tables’.

Almost all adults (97%) reported they were at least aware of AI; however, there was a wide variation in the extent of this by age group. Of those aged 16 to 19, 84% reported that they knew either “a lot” or “a fair amount” about AI prior to their participation in the survey. This percentage decreases gradually with age but remains at over 50% until the 55 to 64 age group (48%). Reported awareness of AI was at its lowest for those in older age groups, with 19% of those aged 75 to 84 and only 9% of those aged 85 or over having heard “a lot” or “a fair amount” about AI.

Figure 6.2 shows the usage of different types of AI by UK adults.

Figure 6.2: AI usage by type of AI technology, UK, 2025/2026 (base: all adults)

Notes: data from figure 6.2 can be found in table E3 of the accompanying ‘DSIT Public Engagement Survey 2025/2026: artificial intelligence tables’.

Overall, 56% of adults had used AI that creates human-like text or speech in response to queries at any point within the last three months, followed by AI-powered digital assistants that could understand natural language (54%) and technology which uses AI to create or edit images, videos or music (32%). Least commonly used were workplace AI systems and robotic technology that used AI to perform its duties, which were used by 19% and 14% of adults in at least the last three months respectively.

Throughout the following set of questions, survey respondents are described as being users (or non-users) of “generative AI”. This was defined as someone who has reported that they had used at least one of the following types of AI at any point within the last three months:

• AI that creates human-like text or speech in response to prompts or queries, such as ChatGPT, Copilot and Gemini 

• technology which uses AI to create or edit an image, video or piece of music 

• workplace AI systems that can perform tasks, make decisions and take action on their own such as setting and prioritising goals or initiating new tasks or workflows 

Those that reported that the only types of AI that they had used in the last three months were robotic technology that uses AI to perform its duties (e.g. robotic vacuums or self-driving features in cars), or digital assistants powered by AI that can recognise voices (e.g. Siri or Alexa) were not classed as users of “generative AI”. 

Figure 6.3 shows the proportion of adults that had used a form of generative AI in the past three months, for different age groups with the UK total included for comparison.

Figure 6.3: Usage of generative AI in the last three months, by age, UK, 2025/2026 (base: all adults)

Notes: data from figure 6.3 can be found in table E5 of the accompanying ‘DSIT Public Engagement Survey 2025/2026: artificial intelligence tables’.

Overall, 59% of adults had used a form of generative AI in the last three months. Usage was 75% or over for adults below the age of 45. After this point, usage decreases with age, down to 18% for those aged 75 to 84 and 8% for those aged 85 or over.

Figure 6.4 shows the proportion of adults that had used a form of generative AI in the past three months, broken down by highest qualification attained.

Figure 6.4: Usage of generative AI in the last three months, by highest qualification, UK, 2025/2026 (base: all adults)

Notes: data from figure 6.4 can be found in table E5 of the accompanying ‘DSIT Public Engagement Survey 2025/2026: artificial intelligence tables’.

Usage of generative AI generally increased with the level of highest qualification, with 75% of those with a degree-level qualification or above having used this type of AI in the last three months, 48% of those with another type of qualification, and 19% for respondents with no qualifications.

Figure 6.5 shows the proportion of adults that had used a form of generative AI in the past three months, for each ITL1 region and country.

Figure 6.5: Usage of generative AI in the last three months, by ITL1 area, UK, 2025/2026 (base: all adults)

Notes: data from figure 6.5 can be found in table E6 of the accompanying ‘DSIT Public Engagement Survey 2025/2026: artificial intelligence tables’.

Use of generative AI varies by area. The highest rates of usage were seen in London, where 69% had reported using generative AI in the last three months. Usage was lowest in North East England (51%).

Users of generative AI who completed the online questionnaire, were then asked a series of questions relating to how they applied these tools and their motivations for doing so. It was not possible to include the questions specifically on use of generative AI within the paper questionnaire. Of generative AI users identified, 97% completed the online questionnaire. Figure 6.6 shows the settings in which adults had used generative AI.

Figure 6.6: Settings in which AI is used, UK, 2025/2026 (base: generative AI users from the online sample)

Notes: data from figure 6.6 can be found in table E7 of the accompanying ‘DSIT Public Engagement Survey 2025/2026: artificial intelligence tables’. This question was not included in the paper questionnaire and so the results are only representative of generative AI users from the online sample.

Overall, 83% of generative AI users used this for personal reasons, 54% for work and 27% for education or study. Naturally among younger age groups there was a higher share of generative AI use for education or study (73% of 16 to 19 year olds and 43% of 20 to 24 year olds).

Figure 6.7 shows the frequency of AI usage by generative AI users within the last three months.

Figure 6.7: Frequency of AI usage in the last three months, UK, 2025/2026 (base: generative AI users from the online sample)

Notes: data from figure 6.7 can be found in table E9 of the accompanying ‘DSIT Public Engagement Survey 2025/2026: artificial intelligence tables’. This question was not included in the paper questionnaire and so the results are only representative of generative AI users from the online sample.

Overall, 75% of generative AI users were using these tools at least once a week, with 35% reporting at least daily usage.

Figure 6.8 shows the types of activities that AI had been used for in the last three months.

Figure 6.8: Activities AI was used for in the last three months, UK, 2025/2026 (base: generative AI users from the online sample)

Notes: data from figure 6.8 can be found in table E11 of the accompanying ‘DSIT Public Engagement Survey 2025/2026: artificial intelligence tables’. This question was not included in the paper questionnaire and so the results are only representative of generative AI users from the online sample.

The most common application of AI was to write or edit documents (58%), followed by work purposes (46%) and editing or creating images, video or sound (40%). Least common applications included support with emotional or relationship issues (10%), applying for jobs (14%) and support with legal or financial issues (16%).

Figure 6.9 shows the different reasons given for using AI tools.

Figure 6.9: Specific reasons for AI usage, UK, 2025/2026 (base: generative AI users from the online sample)

Notes: data from figure 6.9 can be found in table E13 of the accompanying ‘DSIT Public Engagement Survey 2025/2026: artificial intelligence tables’. This question was not included in the paper questionnaire and so the results are only representative of generative AI users from the online sample.

Overall, 65% of generative AI users used it as a way to save time, 55% used AI to help provide ideas or inspiration and 48% used AI to help summarise or explain something. The least common motivation was for companionship (3%).

Figure 6.10 shows the self-reported confidence of generative AI users in applying AI tools, ranging from being able to use advanced features across a range of tools, to having limited experience. The UK total is included for comparison.

Figure 6.10: Self-assessment of AI skills, by age, UK, 2025/2026 (base: generative AI users from the online sample)

Notes: data from figure 6.10 can be found in table E15 of the accompanying ‘DSIT Public Engagement Survey 2025/2026: artificial intelligence tables’. This question was not included in the paper questionnaire and so the results are only representative of generative AI users from the online sample. The 85 or over age group has been excluded from this chart as estimates for some response options have been suppressed due to low statistical reliability. Partial estimates are available in the accompanying data tables.

Of users aged 16 to 19, 79% reported that they could at least use common AI tools efficiently, with 30% of those reporting that they could use advanced features across a wide range of tools. The percentage of users that felt that they could at least use common AI tools efficiently steadily decreases with age to only 19% for users aged 75 to 84, with just 1% reporting confidence with using advanced features across a range of tools.

Survey participants that had not used generative AI in the last three months, were asked about their reasons for not using AI. Figure 6.11 shows the different reasons why non-users of generative AI had not used AI. This question was only asked in the online version of the questionnaire. Of non-users of generative AI identified, 86% are represented by the online questionnaire.

Figure 6.11: Reasons for not using AI, UK, 2025/2026 (base: non-users of generative AI from the online sample)

Notes: data from figure 6.11 can be found in table E17 of the accompanying ‘DSIT Public Engagement Survey 2025/2026: artificial intelligence tables’. This question was not included in the paper questionnaire and so the results are only representative of non-users of generative AI from the online sample.

The most common reason overall was a preference to do things without using AI (44%). 33% were worried about privacy or security when using AI tools, 32% felt that they did not know enough about AI to use it and 31% were worried that it might not be accurate.

6.2 Perceptions of AI 

All survey participants were then asked a series of questions about their perceptions of AI more generally regardless of whether they had used AI. Figure 6.12 shows the perceived societal benefits of people using AI.

Figure 6.12: Perceived benefits of AI, UK, 2025/2026 (base: all adults)

Notes: data from figure 6.12 can be found in table E19 of the accompanying ‘DSIT Public Engagement Survey 2025/2026: artificial intelligence tables’.

Easier access to information or advice (47%), freeing up people’s time (45%) and improved efficiency in workplaces (40%) were the most common perceived benefits overall. A small proportion of adults did not think that there were any benefits of people using AI for society (11%).

Table 6.1 shows the most commonly reported perceived societal benefit of people using AI for each age group.

Table 6.1: Most common perceived benefit of AI, by age, UK, 2025/2026 (base: all adults)

Age group	Most common perceived benefit	Percentage
16 to 19	Easier access to information or advice	53
20 to 24	Easier access to information or advice	48
25 to 34	Frees up people’s time	56
35 to 44	Frees up people’s time	58
45 to 54	Frees up people’s time	54
55 to 64	Easier access to information or advice	47
65 to 74	Helps tackle crime (e.g. facial recognition technology)	38
75 to 84	Improved diagnosis and treatment of medical conditions	39
85 or over	Don’t know	37
Prefer not to say	Easier access to information or advice	32

Notes: data from table 6.1 can be found in table E19 of the accompanying ‘DSIT Public Engagement Survey 2025/2026: artificial intelligence tables’.

The most common perceived benefit of AI for adults aged 16 to 19 (53%), 20 to 24 (48%) and 55 to 64 (47%) was easier access to information or advice. For those 25 to 34 (56%), 35 to 44 (58%) and 45 to 54 (54%), the most common perceived benefit was freeing up people’s time. Adults aged 65 to 74 (38%) most commonly believed that helping tackle crime was a benefit to society of using AI. Adults aged 75 to 84 (39%) most frequently believed that AI use would provide societal benefit by improving diagnosis and treatment of medical conditions.

Respondents were also asked about perceived risks from AI use which are summarised in figure 6.13. 

Figure 6.13: Perceived risks of AI, UK, 2025/2026 (base: all adults)

Notes: data from figure 6.13 can be found in table E21 of the accompanying ‘DSIT Public Engagement Survey 2025/2026: artificial intelligence tables’.

The most common perceived risk of AI use was that information may not be accurate (74%), as well as concerns about privacy and security (69%), the risk of spreading false or misleading information (64%), loss of jobs due to AI (63%), the risk of spreading intentionally misleading images or videos (62%) and the negative impact of people’s ability to think creatively or critically (57%). Only 3% of adults had no concerns at all. Concerns that information may not be accurate was the most common perceived risk within most age groups.

Adults were asked to rate the overall impact of AI on society. These perceptions were rated on a scale from 1 (a very negative impact) to 10 (a very positive impact). Figure 6.14 shows the perceived net impact of AI on society. 

Figure 6.14: Perceived net impact of AI on society, UK, 2025/2026 (base: all adults)

Notes: data from figure 6.14 can be found in table E23 of the accompanying ‘DSIT Public Engagement Survey 2025/2026: artificial intelligence tables’.

Overall, 8% of adults perceived the impacts of AI on society as very negative (a score of 1), with 2% perceiving impacts as very positive (a score of 10). Around half of individuals (58%) were between 4 and 7 on the scale, with the largest proportion of individuals (20%) scoring 5.

Adults were asked to what extent they agreed or disagreed that government legislation was needed in various contexts regarding AI. Figure 6.15 shows adults’ attitudes towards government legislation for AI. 

Figure 6.15: Attitudes towards the extent to which government legislation for AI is needed, UK, 2025/2026 (base: all adults)

Notes: data from figure 6.15 can be found in table E25 of the accompanying ‘DSIT Public Engagement Survey 2025/2026: artificial intelligence tables’.

At least two thirds of adults considered that government legislation was needed to mitigate each of the four risks presented to them. Most adults considered that legislation was needed to prevent AI from replacing jobs (50% strongly agreed, 21% tended to agree), improve transparency about AI system development (55% strongly agreed, 23% tended to agree), reduce bias in AI systems (49% strongly agreed, 22% tended to agree) and make AI systems safer (59% strongly agreed, 21% tended to agree).

Figure 6.16 shows adults’ level of concern about harmful AI-generated content by age, with the UK total included for comparison.

Figure 6.16: Attitudes towards harmful AI-generated content, by age, UK, 2025/2026 (base: all adults)

Notes: data from figure 6.16 can be found in table E27 of the accompanying ‘DSIT Public Engagement Survey 2025/2026: artificial intelligence tables’.

Overall, most adults in the UK were concerned with harmful AI-generated content (57% were very concerned, 31% were somewhat concerned). This strong concern was seen across all age groups. Whilst those aged 85 or over were the least concerned when compared to other age groups, most of this age group were still concerned (52% very concerned, 23% somewhat concerned). 16 to 19 year olds were the least likely to be very concerned of any age group (38% very concerned, 39% somewhat concerned).

1. It is assumed that 8% of ‘small user’ PAF addresses in the UK are non-residential. This is adjusted for in calculating response rates. ↩

2. The individual response rate is calculated assuming an average of 1.89 adults (16+) per household. ↩