go to the original language article
BRUSSELS – Bulgaria has everything necessary to start creating artificial intelligence (AI) models for individual businesses to achieve efficiency. This was stated in an interview with BTA by Eva Maydel, a member of the European Parliament from the European People’s Party group and co-rapporteur of the Legislative Act on Artificial Intelligence.
The aim of the regulation to establish harmonized rules regarding artificial intelligence, which was approved in the European Parliament in March last year, is to ensure safety and respect for fundamental rights while promoting innovation. After passing a final check and being approved by the European Council, the text officially came into force on August 1 of this year. Although the legislative act is already in force, companies in the European Union (EU) still have time to fully prepare, as it becomes fully applicable 24 months after approval – that is, in August 2026, when the final requirements concerning high-risk AI systems come into effect, Maydel added.
According to the MEP, the opportunities for Bulgarian companies, startups, and research institutes are significant in terms of using the law as a chance for growth and innovation.
She explained more about the so-called AI factories being built in the European Union (EU). “One of these factories is in Sofia. Practically every business can, with the help of the AI factory and the INSAIT institute, develop a model that is powered by the necessary information, a niche model aimed at specific business processes,” Maydel emphasized.
She added that for developing artificial intelligence in less risky categories, the requirements are lower. The list of high-risk systems includes those used, for example, in diagnosing diseases, the autopilot systems of cars, as well as biometric identification of individuals, as such systems can be involved in criminal activity. They are required to meet strict requirements before gaining access to the EU market, such as undergoing specific and rigorous testing, transparency, and human oversight.
According to Eva Maydel, the biggest challenge in implementing the Legislative Act on Artificial Intelligence is how companies that want to develop AI in these areas classified as risky are prepared to apply the law.
Daily, companies signal that they want a postponement of the entry into force of at least some parts of this legislation, the MEP stated.
Maydel sees a problem in the way the European Commission presents the law to businesses. “The EC did not provide sufficient and timely guidance on how to comply with the provisions of this law, and this creates great dissatisfaction. There is increasing talk that actually with this law the EU wants to stop the development of AI in Europe, while the idea was exactly the opposite. Everyone wanted to create an atmosphere of trust that would help AI in Europe,” said Maydel.
In addition to the economic benefits and challenges, Maydel also pointed out the impact of artificial intelligence on critical thinking in people and its effect on education. In my observations, the use of applications like ChatGPT at the moment does not pose the greatest risks. The risks from uncontrolled use of social networks by children and generative AI are much greater, she stated. (26.11.2025)