Democrat enters the debate of the European dilemma on AI: regulate without stifling innovation

The medium participates in the forum "The Connected Europe", organized by the association Team Europe, to address the challenge of building a European model of Artificial Intelligence capable of competing with the United States and China without renouncing community rights and values.

6 minutes

Captura de Pantalla 2026 08 26 a las 23.55.19

Captura de Pantalla 2026 08 26 a las 23.55.19

Add DEMÓCRATA to Google

Ask FREN

Published

6 minutes

Most read

How to harness the potential of Artificial Intelligence without losing control over it? It is the question that, with different nuances, is repeated daily in offices in Brussels, among officials, advisors, legislators, companies, and consultants. A question that, at times, is approached almost with the naivety of someone who trusts they can tame a technological revolution with the same instruction manual used to respond to the challenges of the past.

The problem is that Artificial Intelligence does not wait. While technological developments accelerate at an unprecedented speed, community negotiations maintain their own timelines, marked by institutional balance, the search for consensus among member states, and the complexity of building common rules for a market of almost 450 million citizens.

In that space between the speed of technology and the rhythm of European politics, nearly three hundred young professionals are gathering this week in Pamplona to try to provide their own answer. Under the motto "Connected Europe", the association Equipo Europa organizes a discussion forum that brings together institutional representatives, experts, companies, and members of civil society with the aim of putting into context, sizing, and shedding some light on some of the main dilemmas currently facing the European Union.

The immediate news often complicates the calm discussion of issues on which Brussels knows that time is not precisely its main ally. The community agenda progresses amid crises, legislative negotiations, political statements, and geopolitical urgencies. But there are debates that require looking up from the short term.

The regulatory future of the digital sector, European technological autonomy, the development of infrastructures, the ability to attract investment and talent, or the impact of Artificial Intelligence on rights and employment are issues that will determine a large part of Europe's position in the coming decades. Therefore, public and private representatives have taken advantage of these days in Pamplona to pause, analyze, and discuss before the frantic pace of the community capital imposes its own logic again.

Democrat, at the center of the European debate on AI

As a result of his experience and his daily work from Brussels, Demócrata has been part of the inaugural day of the forum, actively participating in the panel dedicated to the advances of Artificial Intelligence and its impact on the rights of European citizens. During his intervention, the correspondent of Demócrata in Brussels, Álvaro Villarroel, raised one of the central questions of the day: the need for Europe to find a balance between protection and competitiveness. "The debate is not just about how much we should regulate, but also how we manage to innovate, compete, and not fall behind."

Villarroel recalled that Artificial Intelligence is, probably, one of the technologies that is transforming society the fastest. Just a few years ago, it seemed like a tool reserved for experts, large companies, or technology laboratories. Today it is part of daily life and is changing the way we work, inform ourselves, research, learn, or even make decisions. Europe has wanted to give its own response to this transformation. It has been a pioneer in establishing a specific framework with the AI Act and has bet on a model based on innovation, but also on the protection of fundamental rights and on trust.

However, the European Union continues to face an uncomfortable reality: the United States and China are advancing with enormous investment capacities, large technology companies, and an increasingly intense race to lead this revolution.

The big question, therefore, is whether Europe can be something more than the great regulator of Artificial Intelligence. If it can also develop its own companies, attract talent, mobilize capital, and create the technologies that will define the future.

A race against the clock

Brussels is aware that it is immersed in a race against time. The stopwatch has been running for a while, and Europe is now trying to develop its own model that reduces some of the technological dependencies that hinder its capacity for growth and limit its decision-making margin.

The Secretary General of the European Data Protection Supervisor, Leonardo Cervera, defended during the debate a clearly European approach to this issue. "What is very important is that we have well-made Artificial Intelligence, because if we are going to do it poorly, it is better not to have it at all."

His approach moves away from any technophobic position. Artificial Intelligence, he argued, can transform entire sectors and bring enormous benefits in areas such as health, productivity, or connectivity. But precisely because of its transformative capacity, it needs principles, rules, and supervision. Cervera used a historical comparison to explain his vision. The first great technology that transformed humanity was fire, but not just any fire: controlled fire. "Always with caution, always with rules, and always supervised," he summarized. A logic that, in his opinion, should also be applied to Artificial Intelligence.

The European official thus defended the need for supervisors capable of putting a stop to those who use these tools for harmful purposes. Among his main concerns, he placed two areas: the development of autonomous weapons and the impact of automation on employment. "If Artificial Intelligence is not regulated and its speed is not moderated, it means automation. Automation means job destruction," he warned, calling for a much broader debate on the relationship between productivity, work, and social welfare.

The issue is not, therefore, solely about determining how many tasks machines will be able to perform, but about deciding which tasks we want them to do and which should remain reserved for human control.

Dependency is also technological

The Renew Europe MEP Oihane Agirregoitea shifted the debate from regulation to a question that Brussels considers increasingly strategic: the ability to produce and scale its own technology. Europe, she recalled, has shown that it can develop norms with global impact. But the crises of recent years have also left a clear lesson: regulating what others produce is not enough to guarantee European autonomy. The shortage of semiconductors, the energy crisis, and the growing geopolitical tensions have made it clear that vulnerability is closely linked to dependency.

For Agirregoitea, Europe must intervene, at least, in four major areas: investment, talent, infrastructure, and market. Financing appears as one of the main challenges. "We have not invested sufficiently in recent years," warned the MEP, who also advocated for facilitating access to capital and the growth of European startups and scale-ups to prevent companies with potential from having to leave the continent to find financing.

Talent is the second element. Europe needs to train, retain, and attract professionals in a context of global competition for highly qualified profiles. The third challenge is infrastructure. Technological infrastructure, computing capacity, cloud, data centers, and energy networks are part of the same equation. Artificial Intelligence needs industrial and technological capacity, but also a huge availability of energy. And there appears one of the great contradictions of the technological race: Europe wants to accelerate the adoption of AI to improve its competitiveness, but that acceleration will also increase energy needs.

During the question-and-answer session, Agirregoitea argued that the deployment of data centers and new AI models must progress in parallel with investments in renewable energies, energy efficiency, and electrical networks. One cannot, he summarized, aspire to massively incorporate AI into industry and everyday life without simultaneously addressing the resources that this model will need to function.

Regulate, simplify, and invest

The head of Innovation at MásOrange, Yaiza Rubio, brought the business perspective to the debate. Europe, she argued, must maintain continuity with its values and rules, but it also needs to ensure that companies can incorporate these tools safely and with a greater capacity to scale. Rubio warned that Artificial Intelligence is already advancing from the experimentation phase to the analysis of specific use cases and towards the automation of increasingly complex systems.

The AI agents, he pointed out, will allow the automation of numerous tasks and will force a reconsideration of the training and incorporation of new professionals into the labor market. The concern particularly affects young people. If certain entry-level functions disappear due to automation, the system will have to find new ways to train those who are graduating from university today and need to gain experience before taking on more complex responsibilities.

The conclusion shared by the participants was that Europe cannot limit itself to choosing between regulation or innovation. It needs both. But it also needs to review how it regulates. Regulatory simplification appears as one of the major priorities for the coming years, especially at a time when Brussels is trying to reduce administrative burdens and make the regulatory framework more accessible for companies and industrial sectors. The final formula proposed during the debate could be summarized in three concepts: smart simplification, targeted investment, and the ability to choose.

More key points, information and questions with FREN

AI-GENERATED CONTENT

What is the current stage of the parliamentary process for the AI Act in the European Union and what are the next steps for its final approval?

As of today (August 2026), the parliamentary process for the European Artificial Intelligence Act (AI Act, Regulation (EU) 2024/1689) is fully completed. The Regulation has already been approved by the European Parliament and the Council, published in the Official Journal of the European Union, and entered into force on August 1, 2024. What is currently underway is not the legislative process, but a gradual implementation phase of its obligations.

1. Status of the legislative procedure

The ordinary procedure for the AI Act was closed in several milestones:

  • Political agreement between Parliament and Council in December 2023, welcomed by the Commission as the conclusion of trilogue negotiations (European Commission note).
  • Formal approval in the European Parliament, which gave its consent to the text of the Regulation in March 2024, as recalled by the Spanish Administration in its summary of the 2024 AI Strategy (link).
  • Final green light from the Council on May 21, 2024, when Member States adopted the Regulation on harmonized AI rules (institutional summary).
  • Publication in the OJ and entry into force on August 1, 2024, twenty days after its publication, as explained by the Commission in “The European Artificial Intelligence Act Enters into Force” (press release) and the Spanish AESIA (Regulation summary).

Therefore, from a parliamentary perspective, there are no pending steps for its “final approval”: the AI Act is already fully effective EU law and directly applicable.

2. Current phase: institutional deployment and phased application

Having concluded the legislative phase, the Union is focused on:

  • The launch of the European Artificial Intelligence Office (AI Office), a Commission body coordinating the application of the Regulation and particularly supervising general-purpose AI models, supported by a European committee and a scientific expert panel, as detailed by the Commission (note).
  • The designation of national competent authorities in each Member State before August 2, 2025, and the deployment of market surveillance, explained in the same note.
  • The development of guidelines, technical standards, and codes of conduct to facilitate compliance, including codes of good practice for general-purpose models and sectoral guides (Commission, Portuguese government).
  • A subsequent reform (“AI Omnibus”) adjusting deadlines and obligations, postponing part of the rules for high-risk systems, covered by the Council and various political chronicles (Council note, analysis in Demócrata).
3. Implementation timeline: what is already in force and what remains

The AI Act is not applied all at once, but through a phased schedule which, after the 2026 reform, can be summarized as follows (synthesis from the Commission, AESIA, and Demócrata analysis):

  • August 1, 2024: entry into force of Regulation (EU) 2024/1689 (AESIA).
  • February 2025 (six months after entry into force): prohibitions of unacceptable risk practices begin to apply (e.g., certain uses of biometric surveillance or subliminal manipulation), according to the Commission (note).
  • August 2, 2025: specific obligations for general-purpose AI models come into force, strengthening transparency and risk management, according to regulatory bodies such as ANACOM (link).
  • August 2, 2026: general application of the Regulation to most systems is reached, including the obligation to inform users when interacting with AI and to technically label synthetic content (deepfakes, generated texts, etc.), as explained in a detailed Demócrata analysis (explanatory article).
  • December 2, 2027: following the 2026 reform, full obligations for “independent” high-risk systems (health, education, employment, access to essential services, migration, justice, border control, etc.) begin to apply (Demócrata).
  • August 2, 2028: obligations for high-risk AI systems integrated into products already covered by other European safety legislation (e.g., certain medical devices or vehicles) are completed, closing the implementation schedule (analysis on deepfakes and transparency).

In summary, the AI Act is no longer “under consideration” by European institutions: it is approved, in force, and in a phase of regulatory deployment and supervision. The “next steps” are not parliamentary but involve consolidating the AI Office, designating and strengthening national authorities, approving harmonized technical standards, and implementing the remaining blocks in 2027 and 2028, especially for high-risk AI.

What are the powers and functions of the European Data Protection Supervisor and what is Leonardo Cervera's professional background?

The European Data Protection Supervisor (EDPS) is the independent authority responsible for ensuring that the institutions and bodies of the European Union respect the fundamental right to personal data protection. Its functions combine supervision, advice, and control in classic privacy areas and new digital frameworks, including artificial intelligence. However, regarding the specific professional background of Leonardo Cervera Navas, the consulted sources do not provide detailed biographical information or an identifiable institutional profile.

Powers and functions of the European Data Protection Supervisor

According to the European Parliament press release on the election of candidates for the position of European Data Protection Supervisor (European Parliament communication), the EDPS is:

  • The EU's independent data protection authority: it is not hierarchically dependent on the Commission, Council, or Parliament, allowing it to exercise its supervisory functions without political instructions.
  • Supervisor of data processing in EU institutions and bodies: it controls how personal data is processed to ensure compliance with current privacy rules (in particular Regulation (EU) 2018/1725, which adapts GDPR principles to European institutions).
  • Advisor on policies and legislation: it advises EU institutions “on all aspects of personal data processing and related policies and legislation.” This includes issuing opinions, recommendations, or observations on European legislative proposals impacting privacy or data use.

Recent sources highlight that the EDPS's powers have expanded with new European digital legislation. The European Parliament note emphasizes that, in recent years, the EDPS's role “has expanded with new digital legislation” and that it is expected to supervise compliance with the Artificial Intelligence Act by EU institutions in their use of AI systems.

This is reinforced in current analyses on the implementation of the European AI Act. An explanatory dossier on the Artificial Intelligence Act —published by Demócrata— recalls that:

  • General compliance monitoring of the AI Act mainly falls to the national competent authorities of Member States.
  • The European Artificial Intelligence Office will have specific powers regarding certain systems and general-purpose models.
  • The European Data Protection Supervisor will act regarding EU institutions, bodies, and agencies, i.e., it will oversee how these entities use AI systems from the perspective of data protection and fundamental rights.

In summary, the main functions of the EDPS can be grouped as follows, based on the consulted sources:

  • Compliance control of data protection rules by all EU institutions, bodies, and agencies (daily processing, large information systems, digital projects, etc.).
  • Consultative function on EU legislative and policy proposals affecting privacy, data economy, digital services, or artificial intelligence.
  • Specific supervision of AI use by EU institutions under the AI Act framework, ensuring systems respect data protection principles and other fundamental rights.
  • Intervention in coordinated supervisory mechanisms alongside national authorities, for example in large IT systems and areas such as Europol or Eurodac, when provided by European legislation.

Available information on Leonardo Cervera Navas

Regarding the second part of your question, specialized tools for institutional profiles and searches in news and documentation consulted have not returned structured biographical results about Leonardo Cervera Navas, nor do they identify him with a specific position within the EDPS in the corpora used.

Specifically:

  • The database of political and institutional stakeholders does not locate any person with that name linked to parliamentary or administrative positions in Spain nor to bodies registered under the name of European Data Protection Supervisor.
  • In textual searches conducted in the press and consulted documentation, no developed biographical profile (education, previous career, appointment dates, etc.) appears under the name “Leonardo Cervera Navas.”

Given this result, and following the criterion of not adding data unsupported by the sources consulted in this session, it is not possible to offer a reliable and detailed reconstruction of his professional background based on these materials. It can be stated that, in the analyzed corpora, his figure does not appear as clearly as that of the EDPS position itself, nor is there an official curriculum that can be precisely cited.

If you need a comprehensive biographical profile of Leonardo Cervera Navas, the most prudent approach would be to consult official sources from the EDPS itself, European institutions, or specific institutional biographies not covered by the databases explored in this research.

What legal requirements must European tech startups meet to access EU public funding in the artificial intelligence sector?

Access for a European tech startup to EU public funding for artificial intelligence projects depends less on a single European “startup law” and more on a set of cross-cutting legal requirements (complying with the EU regulatory framework) and specific eligibility criteria of each program (Horizon Europe, EIC, Digital Europe, structural funds, etc.).

Based on information from the European Commission about the AI innovation package, the European Innovation Council (EIC), and recent initiatives on digital sovereignty and AI gigafactories, a common scheme of conditions can be extracted.

1. Basic company requirements
  • Legal personality and establishment in the EU or associated country. Horizon Europe and EIC calls open the door to companies from Member States and countries associated with the program. In practice, the startup must be legally incorporated in one of those countries and be able to prove it (commercial registry, tax ID, etc.).
  • Fit within the SME/startup category. Programs focused on emerging companies (such as the EIC Accelerator or some Digital Europe lines) target companies that meet the European SME definition and are in early or scaling phases. The Commission itself emphasizes that the EIC Accelerator targets “Europe’s highest-potential startups and SMEs” and that STEP Scale-Up funds companies scaling strategic technologies.
  • Not being in crisis or sanctioned. In nationally co-financed schemes with structural funds (e.g., FEDER aid managed by Red.es to extend AI among SMEs and startups), it is required that the company is not in crisis and is up to date with tax and social security obligations. This criterion is replicated, with nuances, in many European calls.
  • Minimum operational capacity. Instruments like EIC STEP Scale-Up or EIC grants require the company to demonstrate execution capacity: team, technical resources, governance, and financial viability to carry out the project, which is then evaluated through the EIC Fund’s “due diligence.”
2. Alignment with the AI and data regulatory framework
  • Compliance with the European AI Regulation (AI Act). The AI regulatory “sandbox” promoted by Spain together with the Commission has been designed precisely to help startups and SMEs adapt their systems to the requirements of the AI Act, especially when developing high-risk systems. The resulting guidelines serve as practical references to demonstrate compliance, increasingly relevant for obtaining public support.
  • Reliable, safe AI respectful of fundamental rights. The Commission’s AI innovation package insists on funding only AI that is “trustworthy and respects EU values and standards.” In practice, projects must:
    • Provide mechanisms for governance and human oversight.
    • Properly manage data quality and governance.
    • Address risks of discrimination, bias, and security.
    These elements usually appear as technical or impact evaluation criteria.
  • Compliance with data protection and cybersecurity. Although each program addresses this with varying detail, AI calls require full respect for the European data protection framework and digital security rules. This is especially scrutinized in sensitive sectors (health, employment, education, mobility).
3. Specific requirements of European AI programs
  • Horizon Europe ("Digital, Industry and Space" cluster). Funds projects combining scientific excellence and economic and social impact. For AI startups:
    • Usually participation in consortia with other companies, universities, or research centers.
    • Projects aligned with priorities such as trustworthy AI, innovative data services, and EU strategic autonomy.
    • Justification of capacity to exploit results and generate innovation in the European market.
  • European Innovation Council Accelerator (EIC Accelerator). According to the Commission, it is the flagship program for deeptech startups, with deep technological evaluation and business analysis:
    • The company must present a highly disruptive innovation with global scaling potential.
    • Pass a competitive evaluation process by experts and, if applicable, due diligence by the EIC Fund.
    • Accept the combination of grant and possible equity investment, with associated corporate obligations.
  • Digital Europe Program. Funds advanced infrastructures and services, such as AI factories and supercomputing gigafactories:
    • The startup can be a direct beneficiary or preferred user of these infrastructures, but in both cases must be established in the EU and comply with trustworthy AI standards.
    • In infrastructure calls (AI Factories, gigafactories), consortia must contribute to European technological sovereignty and value chains with relevant European components.
  • Structural and investment funds (such as FEDER). Although managed at national or regional level, they follow common EU rules:
    • FEDER co-financed AI aids (such as the Spanish 40 million line to extend AI to startups and SMEs) require the company to meet general beneficiary requirements for EU funds and align with the program’s objective (AI use cases applied to economic activity).
4. Trends in corporate harmonization and facilitation

Meanwhile, the EU is processing the so-called “Regime 28” and the figure of the Unified European Company (S.EU), designed to offer a harmonized corporate framework, digital by default, and with incorporation in 48 hours. Although not a legal requirement to access funds, they aim to reduce legal fragmentation, facilitate cross-border operations, and make it easier for a startup to operate and finance itself in several Member States, which is very relevant for European AI projects.

5. Practical checklist for an AI startup
  • Verify that the company is legally incorporated and registered in the EU or associated country and fits the SME/startup category required by the program.
  • Demonstrate minimum solvency and operational capacity (team, accounts, governance).
  • Have a clear compliance strategy for the AI Act, data protection, and cybersecurity, especially if the system is high-risk.
  • Align the project with official priorities (trustworthy AI, digital sovereignty, industrial applications, social impact) described by the European Commission.
  • Prepare documentation on business model, impact, and scalability, fundamental in the EIC Accelerator and investment instruments like InvestAI or Scaleup Europe Fund.
  • Always review the specific call, where eligibility requirements, exclusions, and evaluation criteria for that particular funding line are detailed.

Play

Test your knowledge with FREN!

How much do you know about this topic? Answer the following 3 questions.

Which forum brought together institutional representatives, experts, and young professionals to debate the regulation of Artificial Intelligence in Europe?

Question 1 of 3

What is one of the main risks mentioned in the debate if artificial intelligence is not properly regulated?

Question 2 of 3

According to MEP Oihane Agirregoitea, in which areas should Europe intervene to strengthen its technological autonomy?

Question 3 of 3

Hola, soy Fren. ¿Cómo te ayudo?