The paths that open up artificial intelligence | The power of business

Public sector

Paul van Branteghem AI expert

He is an advisor to INECO, a Spanish public transport and infrastructure engineering company reporting to the Ministry of Transport and Sustainable Mobility.

Neural trust

Alexandre Domingo Co-founder

NeuralTrust is a “start-up” dedicated to security solutions in generative AI tools, founded in 2022 in Barcelona

Although artificial intelligence (AI) has been discreetly present in our daily lives since the dawn of privacy (think face-recognizing digital cameras or virtual assistants), it has seen massive growth in recent years with the emergence of generative AI. Most adults turn to it to solve doubts, prepare for exams, plan routines, generate images, write texts, translate them or simply bother them. The tool that was not possible was ChatGPT, a type of chatbot who answers questions, writes texts, translates them and creates other content such as images and codes. It is estimated that 10% of the world’s population uses this program from the American company OpenAI in their personal or professional life. And that’s just the beginning, says Alejandro Domingo, co-founder of Neural Trust, a company to start up specialized in cybersecurity applied to artificial intelligence: “The most powerful technology known to humanity promises to have a brutal impact on our lives,” declares this expert in the video which opens this text, the third chapter of the series The power of businessof Banco Sabadell.

“It is a spectacular moment to be alive and participate in this unprecedented transformation,” said Domingo during the conversation with Paul van Branteghem, specialist and advisor to the Secretary of State for Digitalization and Artificial Intelligence of the Spanish Government (SEDIA), on the impact of AI on business and society. In 2022, there were 2,800 companies in Spain dedicated to the development of different aspects of this innovation. “His success is totally transversal, whether it is applied to the bench or to the defense, it can affect all sectors,” said Van Branteghem. I think it should be emphasized that Spain is among the 15 countries in the world that make the most private investments in AI projects in multiple areas.

We analyze how this general-use technology generates a global impact similar to that produced by the arrival of electricity or the irruption of the Internet.

AI is a driver of the Spanish economy, particularly as an accelerator of productivity and digital transformation. The digital sector it encompasses reached the equivalent of 26% of national GDP last year, according to the report. Digital economy in Spain 2024prepared by the Asociación Española de Economía Digital (Adigital).

There are more than 2,800 AI-native companies in Spain, employing 109,200 people. (1.16% of total employment) in 2022, a figure which grows at an annual rate of 10%, i.e. points more than the rest. In addition, these media companies are more profitable than the rest of Spanish companies.

Spanish companies are adopting AI unevenly. However, the percentage of companies with more than 249 employees was used in 2023, even if this percentage decreases considerably in smaller ones. The most common use is text analytics and the most common reason for not using it, arguing for businesses of all sizes, and lack of knowledge on the subject.

The degree of adoption also depends on economic activity. The sector where AI is most established is that of information and communication, followed by that which develops technologies both in scientific and technical fields and professional activities. In glue, construction, transport and storage of products and small commerce.

Spain has prestigious research centers. Barcelona’s National Supercomputer Center, Marenostrum, has been chosen by the European Union (EU) as one of its 15 AI development centers, known as AIFactories. Moreover, it ranks 14th in the Top 500 supercomputers in the world. Santiago de Compostela will house a second factory, dedicated to health innovation, so that Spain joins Poland and Germany as the only territories in the European network with supercomputing centers.

An unprecedented public investment. Since 2020, the Government has allocated 600 million euros to AI projects. In 2024, we added 1.5 billion additional revenues from the European NextGeneration plan (a program aimed at mitigating the effects of the pandemic) which will be used to boost Spanish supercomputing centers. All resources and programs are channeled and coordinated through the National Artificial Intelligence Strategy (ENIA).

Normative leadership. Spain is a proactive country when it comes to adopting AI and its regulation, says Pompeu Casanovas, research professor at the Artificial Intelligence Research Institute (IIIA) of the Higher Council for Scientific Research (CSIC): “The government was the first in the Union to create a workshop to guarantee its ethical and safe use based on European regulations: the Spanish Artificial Intelligence Surveillance Agency (AESIA),” explains the expert.

Solid and pioneering regulatory framework. The EU created the AI ​​Law, the world’s first legal framework that sets standards for these technological systems in an effort to mitigate risks, Casanovas explains: “Prohibits uses such as Chinese social rating systems (which measure the behavior of individuals and companies by points, to reward good conduct, such as paying bills and meeting standards) and penalizes bagging).

Wide availability of fiber optics and advanced connectivity. AI needs high-capacity networks and Spain must excel in digital infrastructure, with wide coverage of high-capacity networks – high-speed optical fiber that reaches businesses and homes – and 5G.

Public alternative to foreign AI. The National Artificial Intelligence Strategy (ENIA) devotes 10 million euros to the creation of an AI model designed in Castellano, Gallego, Europe and Catalonia. A sort of public and open ChatGPT so that Spain does not depend on foreign private companies. You allowed me to create a chatbot internal to the Tax Administration and a consultation system in the public health system for the early diagnosis of heart failure.

Inequalities between regions. The adoption of AI by companies varies between autonomous communities, which creates a digital divide within the Spanish corporate fabric. Madrid and Catalonia are at the forefront, while the Canaries and the autonomous cities of Ceuta and Melilla are close behind.

Guarantee of ethical governance. The main corporate right in relation to AI is to ensure that it is implemented responsibly, safely and fairly. Decisions relating to its development may be in the hands of large companies, usually foreign, so the challenge of Administraciones is to guarantee that:

• Transparent: automated decisions are verifiable and explainable.
• Secure: which protects user privacy.
• Simply: that it is not conditioned by errors which distort the interpretation of the data or results. This information can come from both how the models are trained and the information they are fed. And above all, that AI serves all citizens.

European regulations, according to Casanovas, from CSIC, guarantee these aspects, but they must be continually updated, allowing innovations to develop at high speed. This, Señala Casanovas, requires great efforts on the part of the Administrations. On the other hand, this regulation imposes numerous limitations on European companies, putting them at a disadvantage compared to their competitors in other regions of the world. This, according to the expert, produces a current of thought that defends deregularization within the European Union, which can mitigate risks.

Meet sustainability. Training AI models, as well as the devices needed to use them, involves an environmental cost: high energy consumption that generates waste and polluting emissions and high water consumption (the data centers that train and run the AI ​​models require large refrigeration systems, and these use enormous amounts of water to maintain the appropriate temperature). A simple request for information from ChatGPT, for example, consumes more electricity than a query on a search engine like Google, according to data from the International Energy Agency (IEA).

Promote knowledge in society. It still exists a general lack of knowledge about how AI works and its applications. There Funcas survey on artificial intelligence By 2024, an estimated 44% of respondents believe they have very limited knowledge about AI, and only 6% expressly have advanced knowledge about it.