News

The Number of AI Companies in Spain Has Doubled in Ten Years, but Over Half Are Concentrated in Madrid and Catalonia

The Ramón Areces Foundation and the Ivie have just released a study on the progress of AI firms in Spain and the adaptation of higher education to the demand for professionals in this field

Spain is among the 15 countries that invest the most private capital in artificial intelligence (AI) projects—$362 million in 2023—and also ranks among the 15 that have founded the most AI companies over the past ten years. The number of AI-native companies, i.e., those created specifically to develop various aspects of AI, has grown by 127% in Spain between 2012 and 2022, although the strongest momentum was recorded before the COVID-19 pandemic. Since then, the number of AI companies has remained stable at over 2,800, employing 109,200 workers. Total employment in these companies has grown at an annual rate of 10%, eight percentage points above the average for Spanish companies as a whole (2%), which shows the importance of the development of these businesses. Although their presence in the Spanish business fabric is still limited (0.36% of companies and 1.16% of employment in 2022), the new book La inteligencia artificial en España: formación y empresas nativas (Artificial Intelligence in Spain: Education and Native Companies”, recently published by the Ramón Areces Foundation and Ivie, highlights the value these companies can bring to the Spanish economy.

In terms of the regional distribution, AI-native companies are notably concentrate in the Community of Madrid, followed by Catalonia, which together account for more than 50% of all AI-native companies in Spain and 74.4% of AI-related employment. Madrid is home to 29.3% of these companies, but its weight is even greater in terms of employment, generating 56.8% of the total jobs. Catalan AI-native companies represent 21.9% of the total, but only 17.6% of employment, as they tend to be smaller in size. These two regions also account for the highest percentage of total companies in the country, each with a similar weight of 18%, compared to the higher representation observed among AI companies. In terms of employment, Madrid and the Basque Country are the regions where the share of AI-related jobs exceeds that of the total workforce. Following these leading regions are the Valencian Community (10.3% of AI-native companies and 5% of employment), Andalusia, the Basque Country, and Galicia. In addition to Madrid and Barcelona, the provinces that stand out in terms of number of companies are: Valencia, Alicante, A Coruña, Málaga and Zaragoza.

The geographical polarization of AI development is also seen in the intensity of AI within a region. Once again, Madrid leads in AI implementation, with 0.57% of its companies being AI-native, compared to the national average of 0.36%. The regions with the lowest AI penetration are: Castilla-La Mancha (0.11%), the Balearic Islands (0.19%), and the Canary Islands (0.20%). The difference in penetration between the leading and trailing regions is fivefold in terms of companies and 16-fold in terms of employment.

AI-native companies are more profitable than the average Spanish company and their employment is growing above the national economic average. Despite these strong results, the study highlights that a quarter of AI companies (24.9%) face difficulties accessing financing and 40% are partially restricted. The authors of the study emphasize that “this is a worrying percentage if we want these companies to scale up and create an AI ecosystem in Spain”.

Spain’s productive specialization, based on more traditional sectors, in which more dynamic, technological, and innovation-based sectors carry less weight, is a hindrance to the development of AI. In Spain, sectors with low exposure to AI have a greater weight in added value than the EU-27 average (26% vs. 19.8%), while sectors most exposed to AI have a significantly lower weight in the Spanish economy (18% vs. 22.9% in the EU). The programming, consulting, and other IT-related activities, and information services sectors account for 50.8% of AI-native companies in Spain. This is followed by the professional, scientific, and technical activities sector, with 19.3% of companies.  AI-native companies also exist in 27 other sectors, though with a residual share and presence of companies.

The report, prepared by Ivie researcher, Juan Fernández de Guevara, and economist, Consuelo Mínguez, in collaboration with Ivie technicians, Rodrigo Aragón, Gema Bravo, and Rubén Baeza, includes a section that analyzes AI adoption among Spain’s largest companies, regardless of the sector. Of the 500 largest companies in Spain, 234 (46.8%), representing 57.5% of employment, are already developing AI. Of these, 28.2% have outsourced AI development to a service provider, 17% are developing it internally, and the remaining 1.6% have created their own companies within the group dedicated to developing AI.

The sectors with the highest concentration of large companies carrying out AI are postal and courier activities (100% of large companies in the sector), programming, consulting, and other IT-related activities (93.1%), transport equipment manufacturing (76.2%), transport and storage (64.9%), and professional, scientific, and technical activities (66.7%). In general, AI penetration among the 500 largest companies is higher than in the economy as a whole, given that these companies have the capacity, data, and human and financial resources to support AI development.

AI in Higher Education

The university system is responding to the AI challenge by increasing the number of degrees related to these technologies. Between the 2015/16 and 2022/23 academic years, the number of AI-related degrees (including bachelor’s, master’s, and doctoral degree programs) nearly tripled, rising by 169%, compared to a 16% increase in total university degrees. This increase has gone hand in hand with student demand. In the 2022/2023 academic year, nearly 29,000 students were enrolled in AI-related degrees, representing a 473% increase compared to 2015/16, compared to an 11% increase in the total university student population over the same period. Those enrolled in AI rose from 0.3% to 1.7% of the total.

AI education is growing rapidly, not only in technological fields, but also across all areas of education, including Social and Legal Sciences. In Engineering and Architecture, 9.7% of degrees include AI-related content and account for 6.2% of total enrollment, followed by Sciences (1.2%), Arts and Humanities—which includes animation and video game development—(1%), and Social and Legal Sciences (0.8%). The latter is experiencing the fastest growth in enrollment with a 157% increase between the 2018/19 and 2022/23 academic years.

The rise of AI-related degrees in Social Sciences and Law has given this branch considerable weight among AI courses: 17.9% of AI degrees and 22.3% of those enrolled are in these fields. Engineering and Architecture continue to be the main focus of AI education, accounting for 70.2% of degrees and 66.5% of enrolled students. However, the study warns that this growth in AI education is not mirrored in STEM (Science, Technology, Engineering, and Mathematics) education overall, where enrollment is growing at a below-average rate. In other words, while the number of degrees and students enrolled in AI is growing rapidly, enrollment in STEM as a whole is declining: “Training related to AI and STEM disciplines should not be seen as separate fields, but rather as complementary, since the latter represent a scientific and technical foundation necessary for the development and application of AI,” the authors note.

Private institutions stand out in AI education, particularly in master’s degrees, compared to public institutions and in terms of enrolled students and graduates, though not in terms of degrees offered. More than half of students enrolled (50.3%) and graduates (57.2%) in AI-related bachelor’s and master’s degrees belong to private institutions. This greater presence of private universities is possibly due to their greater response to the demand for AI training and their lower penetration in other non-AI degrees, which are typically offered by public universities.

In the field of education, AI shows once again a high degree of geographical concentration. The Community of Madrid offers 30.8% of all AI-related degrees. Catalonia ranks second with the most AI degrees (18%), but stands out in terms of the number of students enrolled and graduates, at over 30%. These are followed by the Valencian Community (10.7% of AI degrees), Castile and León (7.1%), Andalusia (6.2%), and the Basque Country (6.2%). Murcia, Cantabria, Castile-La Mancha, and Extremadura are the regions with the lowest penetration of AI, where those enrolled in AI-related studies do not reach 0.5% of the region’s total.