Artificial intelligence in UK businesses: 2023 to 2026

• The self-reported use of artificial intelligence (AI) in UK businesses with 10 or more employees has increased from around 12% to around 35% since late 2023, with larger firms more likely to have adopted AI.

• AI adoption to date has been relatively shallow among UK businesses with 10 or more employees, with the average number of AI technologies used per adopting business rising only modestly, from around 1.4 to around 1.6 since late 2023.

• AI use varies across industries, with over half of businesses in information and communication (58%) reporting using AI, compared with much lower levels in construction (13%).

• Improving business operations is the most common use of AI, reported by over 60% of larger businesses; however, this has not yet translated into widespread changes in overall workforce headcount.

• Over half of businesses report effects on creative or design roles when using visual content AI technologies, while a similar proportion report effects on administrative or clerical roles when using AI image processing.

The increasing use of artificial intelligence (AI) across UK businesses has led to interest in understanding how it is being adopted, what it is used for in practice, what barriers businesses face in AI adoption and use, and its potential impacts on business activity and the labour market. This article brings together new insight on AI while highlighting some potential topics for future investigation.

Measurement of AI use remains an evolving area in the UK and internationally. National statistical institutes are continuing to develop approaches to capture the use of AI in businesses. For example, in the United States, the Business Trends and Outlook Survey collects data on business use of AI and its effects on production and labour, while Eurostat measures the adoption of AI technologies among enterprises across European countries.

In the UK, the Business Insights and Conditions Survey is currently the most comprehensive business survey collecting timely information on AI from a business perspective. This contributes to a wider international effort to better understand how AI is being used.

Trends in artificial intelligence adoption

Research by the Federal Reserve Bank of St. Louis combines internationally comparable worker and firm surveys to examine differences in artificial intelligence (AI) adoption rates across countries, as described in their Mind the Gap: AI Adoption in Europe and the United States report. Their analysis shows that adoption rates were higher in the United States (US) than in Europe, with 43% of US workers reporting use of generative AI for their job in early 2026. Among European countries, the UK is one of the higher adopters. Around 36% of UK workers report using AI, compared with lower rates in countries such as France, Germany and Italy. Adoption in the UK appears to be relatively high within Europe, while still below levels observed in the US.

Evidence from firm-level surveys suggests that estimates of AI adoption can vary substantially depending on the measurement approach. For example, the National Bureau of Economic reports much higher rates of firm adoption, based on surveys of senior business executives, as outlined in their Firm Data on AI working paper. Around 78% of US firms and 71% of UK firms have reported use of at least one AI technology.

These differences, in part, reflect how adoption is measured. Lower estimates tend to capture direct, bottom-up use (for example, by workers or specific business functions). Higher estimates reflect a broader, top-down view of whether firms use any AI technology across their operations. Estimates may also vary depending on the size of businesses included, as larger firms generally report higher rates of AI adoption.

Figure 1 shows that use of AI among UK businesses has increased steadily since late 2023. Since September 2023, the proportion of businesses (with 10 or more employees) reporting use of at least one AI technology has increased to just under three times its initial level, rising from around 12% to around 35% of businesses. AI use varies by business size, with 28% of businesses with 0 to 9 employees reporting use of at least one AI technology, compared with 49% of those with 250 or more employees.

Depth of adoption of artificial intelligence technologies

One proxy indicator of AI intensity that we can track is the number of AI technologies used per business. Since September 2023, the average number of AI technologies used per business has increased only modestly, rising from around 1.4 to 1.6 (Figure 2). This would imply relatively limited transformative impacts to date for most AI-adopting firms.

By type of AI technology, large language models were the most widely used among businesses (with 10 or more employees) in June 2026 (18%); this is followed by visual content creation (16%), data processing using machine learning (12%), and image processing using machine learning (6%). Robotics (2%) and other AI technologies (2%) were the least commonly used.

However, this is not a perfect proxy of AI intensity. Not all technologies are relevant to all businesses, and we also want to capture how extensively a particular technology is used across a business to provide further insight on AI intensity. New questions on the extent of AI use and the share of employees using AI were added to the most recent Business Insights and Conditions Survey (BICS) wave.

Among businesses (with 10 or more employees) that report using at least one AI technology in their business operations, only 10% report using AI extensively. This varies by business size, with 17% of businesses with 0 to 9 employees reporting extensive use, compared with 9% of those with 250 or more employees.

A similar pattern is observed for the share of the workforce using AI, with 15% of businesses (with 10 or more employees) reporting that more than half of their employees use AI as part of their daily work.

Larger businesses are much more likely to report being unsure about the extent of AI use and the share of employees using AI. This may reflect differences in how easily these measures can be assessed. As these survey questions have only recently been introduced, these results should be interpreted as early indications; we expect future waves to provide further clarity.

Variation by sector

Levels of AI use vary notably across industries, in terms of overall adoption and the intensity of use (Figure 3). Industries that are considered more digitally intensive or knowledge-based tend to show both higher adoption and greater intensity of use.

Industries such as information and communication, and private sector provision of education, show the highest levels of AI use. Almost three-fifths of businesses in the information and communication industry reported use of at least one AI technology. In these industries, a relatively large proportion of businesses report using multiple AI technologies, indicating more intensive adoption. In contrast, industries such as construction show lower levels of AI use overall (13%).

Comparing employee use and business use

Reported use of AI is higher among employees than among businesses. AI adoption appears to be more widespread at the individual level (Figure 4). Over half of employees (55%) report using AI for work or education, compared with around a third (35%) of businesses reporting use of at least one AI technology.

This might reflect the way AI is being adopted within organisations. Official business measures capture whether firms report using AI technologies. However, individual employees may use AI tools independently, as part of their day-to-day work activities. This might also reflect some of the measurement uncertainties in AI, including how much of AI activity is included as part of the production boundary.

The business measure reflects use of specific AI technologies within firms. The employee measure captures self-reported AI use in work or education, which may include more informal or ad hoc usage.

How business use artificial intelligence

Understanding how artificial intelligence (AI) is used within businesses is important when assessing its potential to improve the efficiency and productivity of existing processes. This might relate to how AI is being integrated into established workflows, production processes or day‑to‑day operations. Examining the purposes for which businesses report using AI provides insight into whether adoption is mainly supporting incremental improvements in performance or more fundamental changes in how firms operate.

Improving business operations is the most reported use of AI across all business size bands (Figure 5). Close to 60% of businesses report using AI for this purpose, with slightly lower rates among smaller businesses. Providing or personalising products or services is the second most common use, though this is more prevalent among smaller businesses, with around 31% of businesses with 0 to 9 employees reporting this use.

Uses such as developing new products or services and exploring new markets are reported by a smaller proportion of businesses across all size bands. However, these activities are more common among smaller businesses than larger ones, suggesting that they may be more likely to use AI in support of expansion- or innovation-focused activities.

This indicates that AI is more commonly used to improve existing operations, rather than to support new market entry or product development, with larger firms tending to focus on efficiency, while smaller businesses use AI more flexibly across activities. This is consistent with external evidence from the Department for Science, Innovation and Technology's AI Adoption Research, published in 2026, which found that increasing efficiency and productivity is the most cited motivation for adopting or expanding the use of AI. This implies that businesses are mainly using AI to enhance existing processes, with more transformative applications being less widespread.

How businesses adopt artificial intelligence

Industries adopt AI through a variety of methods. This includes free-to-use software, purchased external software, ready-to-use, in-house development, and outsourced providers.

Figure 6 shows that businesses in manufacturing, wholesale and retail, accommodation and food services activities, and administrative and support service activities most commonly report using free-to-use software. However, businesses in construction, information and communication, and professional, scientific and technical activities are more likely to access AI through the purchase of external software or ready-to-use services.

The way businesses adopt AI may have implications for how these are recorded in gross domestic product (GDP), particularly in how we might capture where there is no market price (free-to-use software) or where AI technology is developed in-house. These are some of the reflections that are being considered in the statistical community. We will engage with other national statistical institutions as part of informing our own AI thematic accounts, which we are looking to develop by March 2027.

Labour market effects

An important area of interest for policymakers is the potential impact of AI on the labour market. This includes whether AI may automate or displace certain tasks, increase demand for higher-skilled labour, or create new tasks within jobs. One indicator that we track is what impact (if any) AI technologies have had on a business's workforce headcount.

Most businesses report that the use of AI has not resulted in a change to their overall workforce headcount so far (Figure 7). Across all size bands, around half of businesses reported that AI had no impact on headcount, with particularly high shares among businesses with 10 to 249 employees.

A smaller proportion of businesses report reductions in headcount. This is most evident among medium-sized businesses, with just under 7% reporting a decrease. Levels of headcount reduction are slightly lower in small businesses.

A notable proportion of businesses report that the impact of AI on workforce headcount is either uncertain or not applicable, suggesting that for many firms the effects are either still emerging or not yet relevant to their operations.

This is consistent with wider evidence on the impact of AI on the workforce. Recent data from the Bank of England's Decision Maker Panel, published in January 2026, shows that for most firms, the use of AI has had no material impact on the number of employees over the past three years. This reinforces the view that, to date, the adoption of AI is not yet associated with widespread changes in workforce size.

There is some evidence suggesting that the purpose for which AI is used is associated with different workforce outcomes (Figure 8). Businesses that report using AI to improve business operations are more likely to report reductions in workforce headcount than those using AI for other purposes. Around 6% of businesses using AI for operational improvements report a decrease in headcount, compared with a smaller proportion using AI for other purposes.

Of businesses using AI to improve business operations, 63% reported no change in worker headcount, while 1% reported increased headcount and 6% reported decreased headcount. These early indications show that the impact on employment remains limited at this stage.

Job roles affected by artificial intelligence adoption

Another potential labour market insight is identifying which roles have been most affected by the adoption of AI technologies. The roles most affected vary depending on the type of AI technology used.

Figure 9 shows that AI used for visual content creation is most strongly associated with creative or design roles, with over half of businesses using visual content creation AI reporting impacts in these functions.

Image processing using machine learning is associated with comparatively high reported impacts across most role types; 53% of businesses using image processing technologies report impacts on administrative or clerical roles and creative or design roles.

In contrast, businesses using machine learning data processing most commonly report impacts on administrative or clerical roles (41%) and data analysis roles (39%).

These findings imply that AI is influencing the distribution of tasks across different roles within businesses. The type of AI technology used appears to play an important role in shaping where these impacts are felt.

Barriers to artificial intelligence adoption

Our previous analysis found that the most commonly reported factors preventing or delaying the adoption of artificial intelligence (AI) technologies were:

• difficulty identifying business use cases

• cost

• a lack of expertise

More information can be found in our Management practices and the adoption of technology and artificial intelligence in UK firms article.

This shows that firms with stronger management practices are more likely to adopt advanced technologies, including AI. This provides some evidence that organisational capability plays an important role in overcoming these barriers.

This is consistent with evidence from the Department for Business and Trade's UK Innovation Survey 2025: report, which found that innovative businesses are more likely to use AI and other production‑enhancing technologies. This indicates that adoption is closely linked to wider firm‑level innovation capability.

We can track if there have been any considerations that have prevented or delayed a business from adopting AI technologies. A substantial proportion of businesses report that they have not experienced such barriers to adopting AI technologies in the last three months (Figure 10); 41% of UK businesses with 10 or more employees report no barriers to adoption.

Among businesses that do report challenges, a lack of expertise is one of the most cited barriers. This is particularly evident among businesses with 100 to 249 employees, where around 18% report that insufficient expertise has delayed adoption. Cost is also a consistently reported constraint across all size bands, affecting around 7% to 14% of businesses. In contrast, factors such as employee concerns, regulatory issues and product availability are less frequently reported barriers.

Figure 10: Over two-fifths of businesses with 10 or more employees across size bands report no barriers to AI adoption

Factors preventing or delaying adoption of artificial intelligence (AI) technologies, by employment size band, UK, June 2026

Notes:

1. Businesses may report more than one barrier to AI adoption.

2. "Other", "Not sure" and "Business has not attempted to use AI" responses are excluded.

Artificial intelligence expertise

Figure 11 shows that businesses (with 10 or more employees) that have cited a lack of AI expertise as a barrier in the last three months are consistently more likely to invest in workforce development. Around 62% report training or retraining existing staff to integrate AI skills, compared with around 26% of businesses reporting no barriers. This reinforces that capability gaps are strongly associated with internal upskilling as the main response.

Differences across other approaches are less pronounced and appear more variable across Business Insights and Conditions Survey (BICS) waves. Businesses reporting no barriers are slightly more likely to automate or replace roles than those citing a lack of expertise. The recruitment of new staff with AI-related skills and outsourcing roles remain relatively uncommon across both groups.

Integrating artificial intelligence skills

Businesses most commonly report integrating AI skills into their workforce through training or retraining existing staff (Figure 12). This approach is consistently the most widely used across all size bands, with around 40% of medium- to large-sized businesses reporting integrating AI skills. Around 10% of businesses report automating or replacing roles with AI technologies, with relatively little variation across size bands.

Recruitment of new staff with AI-related skills is more common among larger businesses. This rises from around 2% among the smallest businesses to 10% among those with 250 or more employees.

Given that training or retraining existing staff is the most reported approach, this appears to show that businesses are mainly adapting to AI through internal workforce development, rather than external hiring or restructuring. This aligns with earlier findings that AI use is more often associated with changes in roles and tasks, rather than large-scale changes in workforce headcount.

However, this remains limited in practice. Only 11% of businesses (with 10 or more employees) reported that more than half of their workforce has received AI‑related training.

Our Business Insights and Conditions Survey (BICS) covers a broad range of UK industries, based on the same sampling frame as the Monthly Business Survey (MBS).

However, some industries are excluded from BICS, such as agriculture, oil and gas extraction, energy generation and supply, public administration and defence, public provision of education and health, and finance and insurance. In Wave 159, BICS achieved a response rate of 26.7%, with 38,637 businesses responding to the survey. For more detail on this, see our Business Insights and Conditions Survey (BICS) quality and methods information.

In BICS, artificial intelligence (AI) is not defined as a single concept. Instead, it is captured through a list of specific technologies, with the headline measure based on whether businesses report using any of them. While this offers a consistent baseline, it treats all reported uses as equivalent and does not distinguish between light use, user‑level use, and more embedded, production‑level adoption.

New questions introduced in Wave 159 of BICS, on the extent of AI use within businesses and the proportion of employees using AI as part of their daily work, will help measure this more directly in future analyses. To aid interpretation, the full questionnaire for each wave is published alongside its data, in our Business Insights and Conditions Survey questions: 15 June 2026 to 28 June 2026 article.

Some estimates in this article exclude businesses with 0 to 9 employees because their relatively large survey weights can have a disproportionate influence on aggregate results. In other cases, all business size bands are included, or results are presented by size band, to preserve data granularity that would otherwise be lost.

Measurement of artificial intelligence (AI) adoption and use across the UK economy remains a work in progress, for which we welcome user feedback.

This includes exploring new measures of how AI is integrated into business processes, including agentic workflows. We could also consider how to better capture AI intensity and some of the economic impacts on the labour market, and on productivity at the firm level.

Our newly developed Linked Employer-Employee Dataset (LEED) provides new opportunities to investigate how AI affects both firms and workers. There are also insights that we are looking to capture from a household perspective on AI, including how and where it is being implemented.

We are working with other national statistical organisations, the AI Economics Institute, and other government departments to inform how we might reflect AI in our macroeconomic statistics. These developments reflect ongoing user interest in understanding the role of AI in the economy, and the need for statistics that keep pace with rapid technological change.

By March 2027, we aim to have identified the priority measurement topics for AI, including a clear recommendation on the feasibility and value of an AI-thematic account.