ICT usage in enterprises 2026

Almost every second enterprise used AI in 2026

Statistical news from Statistics Sweden 2026-11-05 8.00

In 2026, 49 percent of enterprises with at least 10 employees used artificial intelligence (AI). This was shown by new statistics from the ICT usage in enterprise survey. The results also showed that data analytics, cloud computing services, and e-commerce continue to be important components of enterprises’ digitalisation.

“The use of AI among Swedish enterprises continued to increase in 2026. Statistics Sweden’s data show that AI has become an established technology in many enterprises and is used across several areas of business operations. At the same time, there are still clear differences between industries and size classes in the use of AI,” says Fredric Nyström, statistician at Statistics Sweden.

In 2026, 49 of enterprises with 10 or more employees used some form of AI technology. Compared with 2025, this represents an increase of 14 percentage points. From 2026 onwards, the survey covers additional industries and is based on a new industrial classification, which may affect comparability with previous years. For more information, see Definitions and Explanations.

The use of AI increased substantially with enterprise size, measured by the number of employees. Among enterprises with 10-49 employees, 45 percent used AI, while the corresponding share was 65 percent among enterprises with 50-249 employees. Among enterprises with 250 or more employees, 85 percent reported using AI.

Among enterprises that used AI in 2026, almost six in ten (58 percent) reported using AI for more than one purpose. The most common purpose was marketing or sales, with 55 percent of AI-using enterprises reporting this use. This was followed by administrative processes and business management, at 52 percent. The least common purpose was logistics, which 7 percent of enterprises reported using AI for.

Marketing or sales were the most common purpose for AI use

Enterprises’ purposes for using AI. Share of AI-using enterpriss with 10 or more employees, 2026.

Enterprises’ use of AI was dominated by closed-source software, that is, software whose source code is not available to the user, such as ChatGPT, Copilot, and Gemini. Among enterprises that reported using AI in 2026, 87 percent indicated this mode of acquisition. The second most common acquisition methods were the use of open-source software, and the use of software or systems develop by an external supplier, both reported by 27 percent of enterprises. Open-source software refers to software whose source code is openly available and can be inspected, modified, and further developed. Examples of open-source AI software include DeepSeek and OpenLLaMa. The least common acquisition method was software developed by the enterprise’s own employees, reported by 24 percent of AI-using enterprises. These results suggest that enterprises primarily rely on ready-made commercial AI solutions, while in-house developed solutions remain less common.

AI use was dominated by acquired closed-source software

Enterprises’ methods of acquiring AI. Share of AI-using enterprises with 10 or more employees, 2026.

Among enterprises that reported not using AI, 14 percent were considering adopting it. Among these enterprises, the most commonly reported obstacle was a lack of relevant expertise within the enterprise, cited by 62 percent. The second most common obstacle was concerns about data protection and privacy breaches, reported by 43 percent. The least commonly reported reasons were ethical considerations and the view that AI was not useful for the enterprise, cited by 27 percent and 26 percent, respectively.

Lack of relevant expertise was the most common barrier to AI adoption

Barriers to AI adoption among enterprises that have considered using AI. Percentage of enterprises with 10 or more employees that considered but decided against using AI, 2026.

Digital tools are becoming more common among Swedish enterprises

In 2026, 78 percent of enterprises with 10 or more employees purchased some form of cloud computing service. This represents an increase of 6 percentage points compared with 2025. Among enterprises using cloud computing services, e-mail services were the most common, used by 88 percent of enterprises. This was followed by file storage (86 percent) and office software (78 percent). The least common use of cloude computing services was for AI technologies used to generate text, images or program code, which was reported by 18 percent of enterprises 

E-mail services were the most commonly purchased cloud computing service

Purchase of cloud computing services by type of use. Share of enterprises with 10 or more employees that used any type of cloud computing service, 2026.

In 2026, 47 percent of enterprises used some form of digital business system or analytics tool. Compared with 2025, this was an increase of 2 percentage points. The most common system was Customer Relationship Management (CRM), used by 35 percent of enterprises. This was followed by Enterprise Resource Planning (ERP), used by 31 percent of enterprises. Business Intelligence (BI) was used by 25 percent of enterprises. Furthermore, 32 percent of enterprises reported electronically sharing data with customers or suppliers in their value chain.

In 2026, 39 percent of enterprises carried out some form of data analytics, which was at the same level as in 2025. The most common approach was for the analyses to be performed by the enterprise’s own employees, reported by 33 percent of enterprises. 15 percent engaged an external provider to carry out data analytics on their behalf. Among enterprises whose employees performed data analytics, the analyses were primarily based on customer and transaction data. Information from social media and public websites was also used as input for these analyses.

Customer and transaction data were the most common source for data analytics

Data source used by enterprises’ employees for data analytics. Share of enterprises with 10 or more employees whose employees perform data analytics, 2026

E-commerce is an important sales channel

For most variables in the survey, the reference period is 2026. However, the survey also includes questions on e-commerce that refer to 2025. The figures presented below therefore relate to 2025.

In 2025, 36 percent of enterprises with 10 or more employees engaged in some form of e-commerce, either through web sales or sales conducted via Electronic Data Interchange (EDI) messages. Web sales were the most common form of e-commerce and were used by 31 percent of enterprises, while 11 percent conducted sales via EDI.

E-commerce accounted for a significant share of total turnover. Among enterprises in the surveyed industries, e-commerce sales represented 26,3 percent of total turnover. Although fewer enterprises used EDI, this sales channel accounted for 15,5 percent of total turnover, compared with 10,8 percent for web sales. This means that the majority of e-commerce turnover originated from EDI-type sales, despite less than one-third of e-commerce-active enterprises using this sales channel.

The survey also asked enterprises to report the percentage distribution of the value of their e-commerce sales to customers in different geographical areas. In 2025, sales to customers in Sweden accounted for 57 percent of the total value of e-commerce sales, followed by 26 percent to customers in other EU countries and 17 percent to customers in countries outside the EU. The results show that the Swedish market is the most important market for enterprises engaged in e-commerce, but that a substantial share of e-commerce sales is also directed towards customers in other countries, both within and outside the EU.

The Swedish market was by far the most important market for enterprises’ e-commerce sales

Distribution of the value of e-commerce sales of goods or services to customers in different geographical areas. Share of total e-commerce sales among enterprises with 10 or more employees, 2025.

In addition to the topics presented here, the 2026 survey also included questions on ICT security, ICT specialists, and ICT and the environment.

Definitions and explanations

Changes to the 2026 survey

From 2026 onwards, the survey also covers enterprises in agriculture, forestry and fishing (SNI 01-03), mining and quarrying (SNI 05-09), and financial and insurance activities (SNI 64-66). At the same time, a new industrial classification, SNI 2025 (based on NACE Rev. 2.1), has been introduced. As a result, both the industry structure and, in some cases, the industry classification of enterprises have changed. The results are therefore not always directly comparable with previous years.

For a more detailed description of these changes and their impact on comparability over time, please refer to the Quality Declaration, Production of Statistics, and Statistics Sweden's webpage on the Swedish Standard Industrial Classification (SNI).

Next publishing will be

2027-11-04

Statistical Database

More information is available in the Statistical Database