The international market for artificial intelligence (AI) will expand by 24 to reach around 5 trillion euros in 2040, far above the current figure of around 210,000 million, according to the projections included in the report "The 2040 technology game plan" prepared by Roland Berger.
According to the consultancy, artificial intelligence, semiconductors and data centers will become the three segments with the largest business volume among the 14 technologies it considers key. Taken together, these three areas will account for more than half of the 16 trillion euros that these activities will add at a global scale by the end of the analyzed period.
In this context, the study estimates that Europe could generate around 3 trillion euros in those 14 technologies, which would represent nearly one fifth of the worldwide market, provided it manages to transform its strength in research into mass industrial production.
The report starts from the examination of more than 400 candidate technologies and selects 30 major trends with the potential to redefine the economy and industry by 2040. The combined business volume of these trends will rise from around 2.5 trillion euros to more than 20 trillion, while their share of global industrial GDP will move from 9% to 37%.
Profitability, the great challenge of AI
The document emphasizes that the challenge of artificial intelligence is not limited to its implementation, but to the ability of companies to translate its use into tangible economic results.
In detail, it notes that 78% of organizations reported using AI in 2024, compared with 55% in the previous year. However, it also indicates that 95% of companies report a zero return in their pilot projects and that only 5% has managed to achieve measurable impact on its profit and loss statement.
For Roland Berger, closing the gap between the high level of adoption and the effective generation of value will be the main business challenge in the next decade. In the European case, the report suggests focusing on specialized applications, such as predictive maintenance, fleet optimization, or industrial process control, leveraging sector expertise and the quality of the available data.
The consulting firm points out that the continent has a talent base comparable to that of the United States, with 128,000 graduates per year in information and communication technologies, compared with 123,000 in the North American country. However, the latter launched 40 relevant AI models in 2024, compared with only three in Europe.
Energy will set the pace of growth
The massive deployment of artificial intelligence will also significantly increase the energy consumption of digital infrastructures. According to the report, AI will move from accounting for approximately one-fifth of data center demand to about 60% in 2035.
"The technological race is increasingly an energy race," said Juan Luis Vílchez, senior partner at Roland Berger in Iberia, who has warned that generation, grids, and storage must grow at the same pace.
Vílchez emphasized that Europe maintains leadership positions in microgrids and distributed generation, and also has a relevant industry in electrical equipment and cooling systems for data centers, although he insisted on the urgency of scaling up these capabilities.
Delays and loss of weight in Europe
The study warns that connecting a new data center to the power grid can take up to seven years in Europe, compared with timelines of between one and three years in the United States. It also projects that the North American country will concentrate about 58% of the world’s installed capacity by 2030, while Europe’s share will fall from 13% to 10%.
Roland Berger’s assessment puts the United States ahead in artificial intelligence markets, data centers, and semiconductors, with China closing the gap at great speed. Europe, for its part, retains leadership positions in microgrids, quantum technologies, carbon capture, utilization and storage, and biomaterials.
The consulting firm proposes three lines of action for the continent: to compete globally at scale in the technologies where it already has strong capabilities; to take advantage of the size of the European market in strategic areas; and to attract foreign companies so that they manufacture or develop their activity within Europe.
This last route is presented as the most feasible alternative in the short term for the production of advanced semiconductors, artificial intelligence, and data centers. Even so, the report also detects opportunities in AI applications for regulated sectors, chip manufacturing equipment, and mature semiconductor nodes.
Alliances to win scale in Spain
"Europe has talent and a significant scientific and industrial base, but market fragmentation and the lack of growth-oriented capital prevent advanced technologies from scaling at the pace of the United States or China," has said the senior partner at Roland Berger in Iberia, Bieito Ledo.
In the case of Spanish companies, Ledo has argued for the convenience of focusing on technologies in which European demand will be decisive, relying on long-term investments and business alliances that make it possible to gain size and competitiveness.