Regulation (EU) 2024/1689 laying down harmonised rules on artificial intelligence (the Artificial Intelligence Act – AI Act) is the first European legislative framework dedicated to the development, placing on the market and use of artificial intelligence systems. The AI Act is being applied in phases. As an EU Regulation, it is directly applicable in all Member States. However, national legislation needs to be adopted to establish the implementation framework at national level.
Implementation of the Artificial Intelligence Act in Romania
The AI Act requires each Member State to designate one or more competent authorities responsible for supervising its application and to establish national rules on the institutional framework and the applicable system of sanctions.
In this context, through a Memorandum of the Romanian Government, ANCOM has been proposed to act as the national market surveillance authority and the national single point of contact, alongside other authorities designated as market surveillance authorities, such as:
- The Financial Supervisory Authority and the National Bank of Romania – for high-risk AI systems in the financial services sector, where the placing on the market, putting into service or use of such systems is directly related to the provision of the respective financial services;
- National Supervisory Authority for Personal Data Processing – for high-risk AI systems in the field of biometrics used for law enforcement, border management and the safeguarding of justice and democracy, as well as for high-risk AI systems used in the areas of migration, asylum, border control management, the administration of justice and democratic processes;
- The authorities designated at national level for the supervision of sectors regulated by Union normative acts. ANCOM is the supervisory authority in the field of electromagnetic compatibility and radio equipment, along with other sectoral authorities, such as the National Authority for Consumer Protection and the Labour Inspectorate.
Currently, the competent authorities designated by the Memorandum are working, together with other sectoral authorities, on the development of the national legislative framework necessary for the implementation of the AI Act, which will establish, among other things:
- the competent market surveillance authorities and their duties;
- cooperation mechanisms between the competent authorities;
- the sanctioning procedure and the system of sanctions applicable for infringements of the AI Act.
Thus, ANCOM or the other competent authorities will be able to verify and sanction the non-compliance with the obligations in the AI Act only after the entry into force of the national implementing legislation currently being drafted.
Phased application of the Artificial Intelligence Act
The Artificial Intelligence Act does not become fully applicable on a single date, but provides for different deadlines for different categories of obligations.
According to the timeline provided by art. 113 of the AI Act, as amended by the Omnibus Regulation in the digital field on AI, published today by the European Commission, it becomes applicable starting with August 2, 2026, with certain exceptions. These exceptions concern, for example, specific obligations relating to high-risk AI systems, which will apply at later dates:
- from 2 December 2027, for high-risk systems listed in Annex III to the AI Act;
- from 2 August 2028, for AI systems classified as high risk pursuant to Article 6(1) and Annex I to the AI Act.
The prohibitions applicable to AI systems posing an unacceptable risk entered into force on 2 February 2025. However, the AI Omnibus Regulation in the field of AI introduces two new additional categories of prohibited practices, applicable from December 2, 2026:
- AI systems capable of generating or manipulating realistic images, video or audio depicting the intimate parts of an identifiable natural person or portraying an identifiable natural person engaging in sexually explicit conduct without that person’s consent;
- AI systems capable of generating or manipulating material constituting child pornography material or child pornography performances.
Transparency obligations for AI-generated content
The obligations that enter into force on 2 August 2026 include those relating to transparency requirements on the use of certain AI systems, laid down in Article 50 of the AI Act. These obligations apply to providers of certain types of AI systems, including providers of general-purpose AI systems, as well as to deployers of such systems.
These provisions establish obligations for providers and deployers of certain AI systems such as those that, among other things:
- interact directly with natural persons, generate synthetic content in audio, image, video or text format;
- generate or manipulate audio, image or video content constituting deepfakes;
- generate or alter published texts aimed at informing the public on matters of public interest.
An exception applies to AI systems, including general-purpose AI systems, that generate synthetic audio, image, video or text content that were placed on the market before 2 August 2026. In such cases, the obligation for providers to ensure that AI-generated content is appropriately marked applies from 2 December 2026.
Under the AI Act, providers are natural or legal persons, including public authorities, who develop or commission the development of an AI system and place it on the market or put it into operation under their own name or trademark, for a fee or free of charge.
Deployers are natural or legal persons, including public authorities, who use an AI system under their authority, except where the AI system is used for a personal, non-professional activity.
These obligations are intended to enable users to recognise when they are interacting with an AI system designed to interact directly with them, or when content has been generated or manipulated using AI.
To support the implementation of these obligations, the European Commission has published the Code of Practice on the Transparency of AI-Generated Content, a voluntary document providing practical recommendations to stakeholders, together with the Guidelines clarifying the scope and application of transparency obligations, including requirements relating to interactions with AI systems and the marking and labelling of AI-generated or AI-manipulated content.
The role of ANCOM
At this stage, ANCOM’s activity aims to participate, alongside other competent sectoral authorities, in the process of preparing the national framework for the application of the AI Act, and to inform the public about the relevant developments in this area.
The Authority recommends that all entities developing, placing on the market or using AI systems to assess their obligations under the AI Act and to continue taking the necessary steps to ensure compliance.
For the latest information on the AI Act, we recommend consulting the dedicated section of the ANCOM website and following the Authority’s official social media channels.