The Flemish government has asked the Department of Economy, Science and Innovation (EWI) to map the degree of adoption, use and expertise of Artificial Intelligence (AI) in Flemish companies. The aim of this AI Barometer is to provide a first baseline measurement to provide an up-to-date monitoring of AI maturity, thresholds and needs and to help evaluate the impact of the Flemish action plan in this area. Future measurement moments will provide the opportunity to acquire a longitudinal overview of the evolution of AI at Flemish companies and to consider this as an international benchmark.
This AI Barometer provides a scientifically based picture of the extent to which Flemish companies integrate data-driven applications in their operations and offerings and is based on two crucial methodological principles. First, a large-scale, random sample (sample number of 14,274 companies, 1,717 usable answers in total) representative of the population of Flemish companies according to company size and sector of activity. Second, a validated measuring instrument in line with similar European questionnaires that can serve as a benchmark.
The main findings of the study are as follows:
The adoption of AI technology among Flemish companies is rather limited. Nearly a quarter (23.2%) of companies use at least one AI technology in their operations: 11.3% use one technology, 6.1% use two technologies, 3.7% use three technologies and 2.1% use four or more AI technologies. This means that 76.8% of Flemish companies do not use any AI technology at the time of the study. Most of these companies have no plans to do so either. The use of AI technologies is mainly a matter for medium-sized (50-249 employees) and large (more than 249 employees) companies.
The lack of knowledge, skills and expertise within the enterprise is the main challenge in implementing AI technology. This applies to both adopters and non-adopters of the technology. In addition, companies have difficulty selecting and recruiting employees with the required knowledge, skills and expertise. Companies also find it difficult to estimate possible AI applications, indicating a lack of knowledge about the applicability of AI technology.