We use cookies to personalise the website and offer you the greatest added value. They are, among other purposes, used to analyse visitor usage in order to improve the website for you. By using this website, you agree to their use. Further information can be found in our data privacy statement.



What is AI Act? What will it change and when will it take effect?

PrintMailRate-it

​​​​​​​​​​​​​by Karolina Sieraczek

8 July​ 2024



TABLE OF CONTENTS​​



What is AI Act?


The AI Act, i.e. the Artificial Intelligence Act, is the world’s first regulation on artificial intelligence that was developed by the European Union authorities.

The regulation aims to ensure compliance of AI systems with fundamental rights and values of the EU, and thus ensure safety while using AI.


what is AI Act


Why does the use of artificial intelligence need to be regulated?


The AI Act responds to the pressing need for regulation of artificial intelligence. Not only politicians but also executives of the major technology companies, e.g. Sam Altman – CEO of OpenAI (responsible for ChatGPT) have pointed to the need for coherent regulations governing the use of artificial intelligence.

Such regulations are crucial for the European economy. The regulation must prevent, or at least, hinder the use of AI for violating safety through misinformation and surveillance. At the same time, it may not impede the development of AI in Europe, as this will negatively affect the competitiveness of European companies in the global technology market.

Which entities are subject to the provisions of the AI Act?


The AI Act applies to:

  1. providers placing on the market or putting into service AI systems in the Union, irrespective of whether those providers are established or located within the Union or in a third country;​
  2. deployers of AI systems that have their place of establishment or are located within the Union;
  3. providers and deployers of AI systems that have their place of establishment or are located in a third country, where the output produced by the AI system is used in the Union.

AI ‘provider’ means an entity that develops this system (or that has an AI system developed) and places it on the market or puts the AI system into service under its own name or trademark, whether for payment or free of charge.

A ‘deployer’ means an entity using an AI system under its authority.

Key provisions of the AI Act


The AI Act distinguishes four risk categories of AI systems:

  • minimum risk systems,
  • limited risk systems,
  • high risk systems,
  • unacceptable risk systems.



risk categories of AI systems


AI systems classified as unacceptable will not be approved for use in the EU as they violate fundamental values, while solutions falling into the other categories will have to meet certain legal and technical requirements.

High risk systems in the AI Act


The AI Act classifies technologies that negatively affect the fundamental rights of the user as high risk systems. These are specifically AI systems that are used:

  • for remote biometric identification of natural persons,
  • for recruitment to educational and vocational training institutions,
  • as safety components in the management and operation of road traffic, or in the supply of water, gas, heating or electricity,
  • for the recruitment of employees, in particular to inform about vacancies, to select job applications, to assess candidates during interviews or tests,
  • to evaluate the eligibility of natural persons for essential public assistance benefits and services, as well as to grant, reduce, revoke, or reclaim such benefits and services;
  • for assessing the risk of a natural person offending or reoffending, or the risks faced by potential victims of crime,
  • to detect the emotional state of natural persons,
  • for assisting a judicial authority in researching and interpreting facts and the law and in applying the law to a concrete set of facts.

The AI Act also calls for a database which will include information about deployed high risk systems. 

Prohibited AI technologies


The AI Act further distinguishes prohibited solutions, i.e. technologies that pose the greatest threat to user safety, i.e.:

  • deploying subliminal techniques beyond a person’s consciousness with the objective, or the effect of materially distorting the behaviour of a person,
  • solutions exploiting any of the vulnerabilities of a natural person or a specific group of persons due to their age, disability or mental disorder,
  • social scoring,
  • biometric identification used in publicly accessible space by law enforcement authorities for the purpose of law enforcement.


When will the AI Act come into effect in Poland?


The European Commission first presented the AI Act in April 2021. In December 2023, the EU Council, the EP and the EC reached agreement on the wording of the Regulation. On 21 May 2024, the AI Act was adopted by the Council of the European Union and will be published in the Official Journal of the EU. It will enter into force twenty days after publication. 

The AI Act will be in force for 24 months after the publication in the Official Journal of the EU. Certain provisions will be effective as soon as 3 months later (provisions regarding the establishment of notifying authorities and notified bodies, and provisions regarding the establishment of the AI Board), or 12 months (penalties). 

Summary

The correct use of the AI Act requires, first and foremost, the proper qualification of the AI system in order to identify the obligations incumbent on suppliers, including to obtain permits or register in databases. Until a relevant practice is developed, this can prove to be quite a difficult task. The risk assessment itself is also essential. 

The AI Act has the status of a regulation which means that these provisions must be applied throughout the EU without the need of being implemented. The administrative fines for non-compliance with the AI Act may be up to 30 million euros or 6% of the total worldwide annual turnover for the preceding financial year, whichever is higher.

As the first regulation of this kind, it is certainly a good start to introducing control over the AI technology used within the EU. The need for further regulations and amendments will arise once the provisions are put into practice.


If you have any questions about artificial intelligence and the AI Act, you are welcome to contact us. At Rödl and Partner, we offer comprehensive legal and technical advice on the use of artificial intelligence in business​.

Contact

Contact Person Picture

Karolina Sieraczek

Attorney at law (Poland)

Associate Partner

Send inquiry

Profile



Deutschland Weltweit Search Menu