The European AI Act: what will it change for companies, citizens, and innovation?

MarGib August 13, 2026

🌐 🇵🇱 Polski · 🇬🇧 EN

The European Union has just introduced the world's first comprehensive regulations on artificial intelligence – the European AI Act. This is a breakthrough step designed to ensure safety, transparency, and trust in AI technology, but it also raises concerns about the competitiveness of European companies against the US and China. What are the main assumptions of the AI Act, what oversight mechanisms does it introduce, and what challenges lie ahead for enterprises and institutions?

For several years, artificial intelligence has been continuously revolutionizing almost every field of life – from medicine and finance to industry and public services. However, along with its growing influence, questions about safety, privacy, and ethics are emerging. The European Union has decided to address these challenges by introducing the world's first comprehensive regulations on AI – the European AI Act. This document is intended to shape the future of AI technology not only in Europe but globally, positioning the EU as a global leader in the field of responsible artificial intelligence development.

The AI Act is not the only element of the European AI strategy, however. It is part of a broader approach that includes funding research, creating trusted AI systems, and promoting ethical standards. What are the main goals and assumptions of the EU's AI policy, what supervisory mechanisms does it introduce, and what challenges does the European economy face in the context of global competition?

The AI Act – what is it and why is it a breakthrough?

The European AI Act is the world's first comprehensive law regulating artificial intelligence. It was proposed by the European Commission in April 2021, and the final text was adopted by the European Parliament in June 2023 and approved by the Council of the EU in February 2024. The Act will enter into force within the next 2–3 years, depending on the risk category of the AI systems.

The main assumption of the AI Act is a risk-based approach, which divides AI systems into four categories:

Unacceptable risk – systems that pose a threat to fundamental human rights, e.g., behavioral manipulation systems or social scoring systems (like those in China). Their use is strictly prohibited.
High risk – systems that may have a serious impact on health, safety, or fundamental human rights, e.g., AI in medicine (disease diagnosis), autonomous vehicles, or recruitment systems. Such systems must undergo a strictly defined conformity assessment and meet a range of technical and ethical requirements.
Limited risk – systems that require a certain level of transparency, e.g., chatbots or emotion recognition systems. The user must be informed that they are communicating with a machine.
Minimal risk – systems that do not pose a threat, e.g., spam filters or recommendation systems. No mandatory regulations are provided for these, although voluntary guidelines regarding responsibility and ethics may be applied.

The AI Act also introduces an obligation to disclose interactions with AI – users must be clearly informed when they are interacting with an artificial intelligence system. Furthermore, high-risk systems will have to be technically documented, and their creators will be held responsible for any damages caused by these systems.

Ethical principles and values behind the European AI strategy

From the very beginning, the European Union has emphasized that its approach to AI must be consistent with democratic values, human rights, and principles of social justice. In 2019, the European Commission published the Ethics Guidelines for Trustworthy AI, which define seven key requirements for AI systems:

Human agency and oversight – humans must have control over decisions made by AI systems, especially in the context of sensitive areas such as medicine or the judiciary.
Technical robustness and safety – AI systems must be reliable, resistant to attacks, and predictable in their operation.
Privacy and data governance – data used to train AI systems must be anonymized and secured, in accordance with GDPR regulations.
Transparency – users must know how the AI system works, what data is being collected, and how decisions are being made.
Diversity, non-discrimination, and fairness – AI systems must not reinforce existing biases or discrimination, e.g., based on gender, race, or ethnic origin.
Societal and environmental well-being – AI should serve the common good, not just commercial interests. The impact on the natural environment must be taken into account.
Accountability – creators and users of AI systems must be responsible for their operation, including for any potential damages caused by these systems.

These principles are not merely abstract guidelines – they are embedded in the AI Act and will be enforced by relevant supervisory authorities. Furthermore, the EU promotes the European AI Pact, a voluntary initiative for companies that want to implement the principles of trustworthy AI in their operations.

What supervisory mechanisms does the EU introduce?

To ensure the effective implementation of the AI Act, the European Union has created a series of supervisory mechanisms and institutions responsible for monitoring the regulations:

European AI Office

This is a new body established by the European Commission in January 2024, whose main task is oversight of the AI Act implementation, cooperation with member states, and the development of technical standards. The AI Office will also monitor progress in the field of AI and issue guidelines for enterprises.

European AI Board

This is an advisory body composed of representatives from member states, experts, and representatives of civil society. Its task is to coordinate actions between EU countries, exchange best practices, and develop standards for AI systems.

National supervisory authorities

Each member state must establish a competent national authority responsible for AI oversight. In Poland, these roles are fulfilled by:

Personal Data Protection Office (UODO) – responsible, among other things, for the compliance of AI systems with GDPR.
President of the Office of Electronic Communications (UKE) – oversees AI systems in the telecommunications sector.

In other EU countries, this role may be performed by, for example, product safety authorities (e.g., in Germany) or cybersecurity agencies.

European Trustworthy AI system

This is a voluntary certification system designed to help companies prove that their AI systems meet European ethical and technical standards. Certification is based on the seven requirements of trustworthy AI (see point above) and is conducted by accredited certification bodies. Although certification is voluntary, companies that obtain it will gain greater trust from customers and business partners.

A pilot certification program was launched in 2022, and full certification is expected to be introduced along with the implementation of the AI Act, i.e., from 2025–2026.

How does the EU fund AI development? Programs and initiatives

The European Union not only regulates AI but also actively supports its development through funding for research and implementation. The most important EU programs are:

Horizon Europe

This is the largest EU research and innovation program for 2021–2027, within which 1.5 billion EUR has been allocated for AI research. The program includes, among others:

Projects regarding machine learning and neural networks.
Research on ethical and responsible AI.
Cooperation between research centers and industry.

Digital Europe Programme

This program, with a budget of 2.5 billion EUR, aims at implementing AI in practice. It includes, among others:

Building European supercomputer centers (eurohpc), which will support the development of advanced AI systems.
Training and development of digital skills, including AI courses for employees and students.
Support for start-ups and SMEs implementing innovative AI-based solutions.

National and regional initiatives

EU countries also actively fund AI development. Examples:

Poland: The "AI for the Economy" program of the National Centre for Research and Development (NCBR) – 100 million PLN for AI projects.
Germany: The "KI-Innovationswettbewerb" program – 500 million EUR for AI innovation.
France: The "France 2030" program – 2.5 billion EUR for AI and future technologies.

What challenges does the European AI strategy face?

Despite ambitious plans, the European AI strategy faces a number of challenges that may hinder its effective implementation:

The issue of competitiveness

The European Union has significantly lower investments in AI than the United States and China. According to a 2023 McKinsey report:

The US invests 2 times more in AI than the EU.
China is implementing innovations faster in the field of AI, especially in sectors such as facial recognition or autonomous vehicles.
Europe is also losing AI talent – according to the European Centre for the Development of Vocational Training (CEDEFOP), the EU has 4 times fewer AI experts than the US.

European industrialists, e.g., Siemens and SAP, also point to excessive bureaucracy associated with the AI Act, which may delay implementations and discourage investment.

Implementation problems and data shortages

For AI systems to work effectively, they need large, high-quality datasets. However, several obstacles are encountered in Europe:

Insufficient data availability – especially in sensitive sectors such as medicine, where patient data is protected by GDPR regulations.
Difficulties in data anonymization – although GDPR imposes an obligation to protect data, in practice, it is often difficult to completely anonymize data without losing its utility.
Lack of uniform standards – EU countries apply different approaches to collecting and sharing data, which hinders international cooperation.

Critical voices regarding the AI Act

There is no shortage of critical voices regarding the AI Act:

Amnesty International points out that the regulations do not go far enough in protecting human rights, especially in the context of mass surveillance (e.g., facial recognition systems used by the police).
Access Now criticizes the AI Act for the lack of sufficient protection against discrimination and bias in AI systems.
Entrepreneurs fear that overly rigorous regulations may limit innovation, especially in the case of small and medium-sized enterprises.

AI and the labor market – revolution or threat?

Artificial intelligence is not only changing the way we work but also creating new challenges for the labor market. According to reports, AI can both replace some job positions and create new ones. What are the opportunities and threats?

On one hand, AI can automate routine and repetitive tasks, such as:

Customer service (chatbots, voice systems).
Data analysis (financial reports, sales forecasts).
Production processes (robotics, machine control).

On the other hand, AI creates new professional opportunities, such as:

AI ethics specialists (people responsible for ensuring that AI systems are consistent with values and regulations).
AI engineers (creators and trainers of artificial intelligence models).
AI cybersecurity experts (people securing AI systems against attacks).

It is worth remembering, however, that AI will not completely replace human labor, especially in areas requiring empathy, creativity, or decision-making in uncertain situations. How, then, to prepare for changes in the labor market? Primarily through continuous learning and developing new competencies.

You can read more about this in our article: Artificial Intelligence and the Labor Market: Revolution, Threats, and New Opportunities.

The future of AI in Europe – opportunities and challenges

The European Union has set itself an ambitious goal: to become a global leader in the field of responsible and ethical AI. Will its strategy succeed? That depends on several key factors:

Advantages of the European approach to AI

Pioneering regulations – the AI Act is the first comprehensive legal act regarding AI in the world, which gives the EU a chance to set global standards.
International cooperation – the EU cooperates with other countries, e.g., Canada and Japan, to develop a common approach to AI.
Funding for research and innovation – programs such as Horizon Europe and the Digital Europe Programme provide funds for AI development.
Ethical approach – the EU emphasizes responsibility, transparency, and the protection of human rights, which can serve as a model for other regions of the world.

Main challenges and threats

Competition with the US and China – investments and the pace of AI implementation in these countries are significantly higher than in Europe.
Bureaucracy and excessive regulation – overly rigorous regulations may limit innovation and discourage investment.
Talent shortage – Europe must attract more AI experts to meet market demand.
Data protection vs. AI development – GDPR and other data protection regulations may hinder the collection and use of data needed to train AI systems.

Despite these challenges, the European AI strategy has enormous potential. If it can be effectively implemented, the EU can become a global leader in the field of trusted, ethical, and responsible AI – which will bring benefits not only to Europe but to the entire world.