As the global landscape of artificial intelligence (AI) regulation evolves, Croatia finds itself at a crossroads. The AI Act, which came into effect at the beginning of August, marks the world’s first significant attempt to place AI under regulatory oversight. However, for now, businesses and developers in Croatia do not need to immediately adjust to these new rules. The real impact will unfold in the coming years, as the technical documents that will shape the regulation are still being crafted.
This phase is crucial, and many experts warn that Croatia cannot afford to be a passive observer. Instead, the country must actively participate in the formation of these rules to ensure that its interests are represented.
Key Involvement Opportunity
The European Commission’s Office for Artificial Intelligence, the body responsible for implementing and overseeing the AI Act, has issued a call to action. It invites companies, individuals, and experts to get involved in creating the Code of Practice for general-purpose AI models. According to Marijana Šarolić Robić, a lawyer and active startup mentor with a special interest in AI, this is a critical moment for stakeholders to influence the development of the regulation.
“The AI Act is a foundational document, but the technical guidelines that follow will be key,” says Šarolić Robić. “It’s crucial for the business community to engage now, while the rules are still being shaped. This is when we can truly influence the outcome.”
The invitation to participate in the Code of Practice development is open until August 25, with the first meetings scheduled for September. These meetings will focus on four key areas: transparency and copyright, systemic risk classification, systemic risk assessment, and mitigation measures. The final version of the Code is expected to be presented in May 2025.
Domestic Participation and Concerns
One of Croatia’s promising AI startups, Omnisearch, is currently exploring how to best engage in this regulatory process. While Omnisearch falls into the low-risk category under the AI Act due to its work with isolated data sets, its co-founder and CEO, Marin Smiljanić, recognizes the broader impact of the regulation on the industry.
“The AI Act aims to prevent dangerous uses of AI, like large-scale facial recognition databases or dystopian social scoring,” Smiljanić explains. “However, there is a concern that the EU’s stringent regulatory approach could be detrimental, particularly for smaller startups. The burden of compliance may be manageable for tech giants like OpenAI, but it could stifle innovation for smaller players.”
Innovation at Risk?
The fear that AI regulation could stifle innovation is shared by many in the Croatian tech community. Šarolić Robić has long warned that over-regulation could kill innovation, and she continues to emphasize the need for balance.
“Regulation is necessary, but it must not suffocate the very innovation it seeks to control,” she cautions. “It’s now a reality we must navigate, and it’s vital that Croatian interests are actively represented in the EU institutions.”
The responsibility for this representation will soon fall to the Directorate for the Digital Economy, which operates within the Ministry of Justice, Administration, and Digital Transformation. As Croatia’s new government forms, there is hope that AI will be recognized as a key area for building competencies and driving initiatives in Brussels.
A Strategic Opportunity
To seize this opportunity, Croatia must define its goals and identify which sectors of AI it wants to promote. However, the national strategy for AI development is still awaiting finalization, leaving the country at a critical juncture.
For Croatia, this is a moment to move from the sidelines to the forefront of AI regulation, ensuring that the nation’s interests are protected and its innovation potential is not stifled by overbearing regulation.
