After publication in the Official Journal of the European Union on July 12, the AI Act (Regulation on Artificial Intelligence, Regulation (EU) 2024/1689) has entered into force today.
An overview of the AI Regulation provisions can be found in our blog article here.
The complete provisions of the AI Regulation will apply from August 2, 2026, but some provisions will apply earlier. We would like to present these in more detail to provide an overview of which provisions will first apply in a few months:
Applicable Provisions from February 2, 2025
According to Art. 113(3)(a) AI Act, Chapters I and II apply from February 2, 2025.
Chapter 1
Chapter 1 regulates:
- the purpose of the AI Regulation (Art. 1 AI Act), namely
“to improve the functioning of the internal market and promote the uptake of human-centric and trustworthy artificial intelligence (AI), while ensuring a high level of protection of health, safety, fundamental rights enshrined in the Charter, including democracy, the rule of law and environmental protection, against the harmful effects of AI systems in the Union, and to support innovation.”
the scope of application (Art. 2 AI Act), namely applicability to
Providers and deployers of AI systems (what matters is whether they have their headquarters in the European Union or put their AI models into service or on the market in the EU, or whether the output is used in the EU);
Importers and distributors of AI systems;
Product manufacturers who place AI systems on the market or put them into service together with their product under their own name or trademark;
Authorized representatives of providers not established in the Union;
Affected persons located in the Union. Exceptions exist, e.g., for private individuals (exclusively personal and non-professional AI use), military and scientific purposes, and for certain AI systems classified as non-risky that are provided under free and open-source licenses.
the definitions (Art. 3 AI Act) that apply to the entire AI Act
AI literacy (Art. 4 AI Act), this article provides that providers and deployers of AI systems must have a sufficient level of AI literacy. The definition of “deployer” according to Art. 3 No. 4 AI Act should be understood quite broadly, namely any “natural or legal person, public authority, agency or other body using an AI system under its authority” (only use within a personal and non-professional activity is excluded, which is already excluded from the scope of the AI Act).
Providers and deployers of AI systems must therefore ensure that their staff and other vicarious agents involved in the use of AI systems have a sufficient level of AI literacy, which depends on
- technical knowledge
- experience
- education and training and
- the context in which the AI systems are to be used
But what is to be understood as AI literacy? This is actually very vague. Art. 3 No. 56 AI Act defines AI literacy as follows:
“skills, knowledge and understanding that allow providers, deployers and affected persons, taking into account their respective rights and obligations in the context of this Regulation, to make an informed deployment of AI systems and to gain awareness about the opportunities and risks of AI and possible harm it can cause.”
How exactly AI literacy is to be conveyed is therefore not defined by the AI Regulation. As shown, it depends primarily on the intended AI use. innFactory AI Consulting GmbH offers corresponding AI training for companies starting in October.
Chapter 2
Chapter 2 concerns prohibited practices in the AI field, which are regulated in Art. 5 AI Regulation. These are AI applications classified as so dangerous that they are not wanted in the EU, such as social scoring or emotion recognition. The fines here are also particularly high: In case of a violation, these can be punished with up to 35 million euros or up to 7% of worldwide annual turnover (whichever is higher).
Applicable Provisions from August 2, 2025
From August 2, the following will apply:
the provisions for establishing the notification procedure for high-risk AI (Chapter III Section 4)
the provisions for general-purpose AI models (Chapter V)
the provisions for establishing governance at EU level in the form of various bodies and at Member State level in the form of a supervisory authority (Chapter VII)
the fine provisions (Chapter XII) with the exception of fines for providers of general-purpose AI systems (Art. 101), whereby these are graduated as follows:
for prohibited AI systems up to 35 million euros or up to 7% of total worldwide annual turnover of the previous financial year, whichever is higher;
outside prohibited AI systems up to 15 million euros or up to 3% of turnover for certain listed violations
for providing false information to notifying bodies or authorities up to 7.5 million euros or up to 1% of annual turnover
the provisions on confidentiality of authorities and bodies provided for under the AI Act (Art. 78)
Of particular relevance from August 2 are the provisions for general-purpose AI models, so-called General-Purpose AI models (GPAI). These only came into the AI Act at a very late stage in the legislative process because ChatGPT had gained such significance by that point. GPAI providers have documentation and transparency obligations in particular (Art. 53 AI Act) and, if classified as an AI model with systemic risk, further obligations, in particular conducting a model evaluation, risk minimization, and an obligation to report serious incidents to authorities.