The AI Act Regulation came into force on 1 August 2024 and aims to provide a legal framework for the development and use of AI technologies in the EU. The regulation introduces basic rules to protect the health, safety and fundamental rights of users of AI systems. It also imposes a number of requirements on suppliers, importers, distributors and users, who have 24 months to comply from the date of entry into force of the regulation. However, some provisions are already in force.

Have a question or need support?

Contact us

We now live at a time when artificial intelligence is not just an invention of the creators of science fiction novels, but our everyday reality, which is an increasing driver of the global economy. It could be said to be a modern-day 'cornerstone', but in the realm of digital innovation, transforming more and more industries of the global economy at an ever-increasing pace and significantly changing our daily lives.

How AI is changing the economy

As recently as 2024, the AI market was valued at USD 184 billion. Compared to 2023, the valuation has increased by almost USD 50 billion. By 2030, it is estimated that the artificial intelligence market will be worth almost USD 827 billion[1].

Now, 77% of companies say they are using AI tools or at least they are exploring their functionality. In turn, 83% of companies assume that AI will be their key development priority in the near term, with 63% of respondents intend to implement AI solutions in their operations within the next 3 years[2]. These statistics and trends paint a vivid picture of the widespread adoption of AI and its huge impact on the economy.

User protection

Interestingly, only 1/3 of consumers are aware of the use of AI in their daily lives, while the actual percentage of AI users is 77%[3]. Given that so few people are aware of contact with artificial intelligence, and that the technology can bring as much good as bad, first world economies have taken to creating regulations to protect the interests of AI users.

At the forefront is the European Union, which has created the first comprehensive act on AI (the so-called AI Act available at the following link: Regulation - UE - 2024/1689 - EN - EUR-Lex). The AI Act is a regulation (i.e. a regulation directly applicable in all EU member states, without the need for transposition steps), which was adopted on 13 March 2024 and came into force only a few months later, i.e. on 1 August 2024.

It aims to create a legal framework for the development and use of AI technology in the EU. The AI Act aims to make AI safe, trustworthy and in line with EU values. The regulation introduces basic principles to protect the health, safety and rights of basic users of AI systems. It applies to a wide range of entities, including AI system providers, importers and users of AI systems operating in the European Economic Area (EEA). However, the regulations also apply to companies outside the EEA that offer their services in Europe.

AI risk categories

The AI Act introduces the categorization of AI systems into several core areas:

  1. AI of unacceptable risk (prohibited AI systems), e.g. social scoring systems that manipulate behaviour or exploit user vulnerabilities, remote biometric identification systems, emotion scoring systems or crime prediction systems.
  2. High-risk AI, e.g. systems for biometric identification and categorisation of individuals, systems for critical infrastructure management, assessment systems in education or employee recruitment and management, systems for access to public services, law enforcement or migration and border control management.
  3. Limited-risk AI, e.g. chatbots, automated search engine suggestions, entertainment applications, data analytics systems, content generators.
  4. Minimum-risk AI, e.g. spam filters, simple recommendation systems, AI-enabled computer games, sentiment analysis systems or tools for automating simple tasks.

Responsibilities and control principles

Depending on the category, there are different requirements and control rules for AI systems. These range from stringent for those categorized as high risk, to minimal for minimum risk AI systems. The basic obligations for the above systems can be summarized as follows:

  1. AI of unacceptable risk (prohibited AI systems): banning the deployment of such systems.
  2. High-risk AI: the need to meet stringent requirements for risk management, transparency and data protection.
  3. Limited-risk AI: the need to ensure that users are informed about the operation of the system and that they are using AI technology.
  4. Minimum risk AI: used without additional legal requirements, although general principles such as ensuring transparency and avoiding discrimination are recommended.

In order to meet these requirements, the AI Act imposes a number of requirements in particular on suppliers, importers, distributors, as well as users of AI systems, including the obligation to carry out security testing and ensure transparency in their operation.

Which regulations are already in force

The regulation provides 24 months from the date of entry into force to comply with most of the requirements. Nevertheless, some provisions have already started to apply! As of 2 February 2025, particularly dangerous systems, i.e. systems that meet the criteria of prohibited AI systems, may no longer be used and marketed. These can include systems that use subliminal, manipulative or misleading techniques, social scoring systems or crime prediction systems.

It is worth keeping an eye on, as the penalties for breaching the above rule can be enormous - up to EUR 35 million or 7% of annual worldwide turnover. They can be targeted at both individuals and legal entities. Importantly, the penalties will take effect slightly later, from 2 August 2025. It is therefore worth looking at the AI systems in use and verifying that they do not fall into the categories of prohibited AI systems to avoid potential penalties.

We cordially invite you to follow our blog, where we will regularly publish articles introducing you to the specifics of this brand new legal regulation and we will inform you about the next key dates related to its implementation. In the articles, we will also not omit the Polish legislation, where work is already underway on a law concerning the issue of AI systems and the implementation of certain obligations imposed on EU member states by the AI Act.

[1] Statista.com, Artificial Intelligence – Worldwide, https://www.statista.com/outlook/tmo/artificial-intelligence/worldwide.

[2] National University, 131 AI Statistics and Trends for 2025, https://www.nu.edu/blog/ai-statistics-trends/.

[3] National University, 131 AI Statistics and Trends for 2025, https://www.nu.edu/blog/ai-statistics-trends/.