At the “Innovate with Data” conference, Aleksandar Marinković, director of ASEE in Serbia, highlighted that the application of artificial intelligence (AI) in local businesses is still in its early stages. Key obstacles to wider implementation include a lack of technology awareness, low levels of digitalization, high implementation costs, and a shortage of qualified personnel.
Marinković referenced EUROSTAT data indicating that only 8% of European companies utilized AI in their operations in 2023, while Serbia lags behind at just 2%. Although the government is proactively supporting AI adoption, the pace of technological development is outstripping the speed of implementation. The availability of skilled workers in the field of AI remains limited, and the educational system is not producing enough qualified professionals.
For small and medium-sized enterprises (SMEs), introducing AI requires substantial resources, including equipment purchases, software development, and employee training. Many companies find it challenging to invest in these technologies without external support or subsidies.
Marinković emphasized that successful AI implementation necessitates prior data consolidation and migration to modern platforms. He pointed out that ASEE spends about 80% of its time on data preparation in countries like the USA, Brazil, and Austria, with only 20% dedicated to AI model implementation. This challenge is not unique to Serbia; it is also faced by companies in Western countries.
He noted that businesses recognizing AI as a crucial value rather than just an enhancement gain a competitive advantage in the market. To reduce the costs of AI adoption for SMEs, Marinković proposed using cloud technologies and AI-as-a-Service models, which eliminate the need for significant investments in internal infrastructure.
Partnerships with larger companies can facilitate cost-sharing and ease the implementation of advanced technologies. Additionally, subsidies and EU funds can help overcome financial barriers, while education will enable companies to recognize the advantages of AI in their sectors.
Marinković believes that AI can improve operational efficiency and speed, enhance customer experiences, facilitate data-driven decision-making, and boost innovation and security.
He provided examples of successful AI applications within ASEE, such as implementing AI models in customer service centers, which improved operations and customer satisfaction. He also noted an increase in the conversion rate for e-commerce clients from 2% to 12%.
Discussing regulations, Marinković underscored the importance of the recently adopted European AI Act, which aims to balance technological innovation with the protection of citizens’ rights. The law ensures safety and fundamental rights while enabling further AI integration, investment in new technologies and the application of innovations.
The law adopts a risk-based approach, meaning different AI applications are subject to varying levels of regulation. This framework penalizes uses of AI with predominantly negative aspects.
Marinković highlighted that the law has extraterritorial applicability, effectively setting a global standard for the ethical use of AI technology, which is significant in a global context.
For further development of the IT industry in Serbia, Marinković stressed the importance of enhancing digital infrastructure, such as the 5G network and supporting education and workforce development. There is a need to adapt educational programs to meet the demands of the growing IT sector and to introduce more courses in programming, artificial intelligence, and cybersecurity. Additionally, expanding retraining programs for those transitioning into the IT field will help include more people and provide support for various IT specialties.
