Artificial intelligence (AI) is rapidly transforming education, significantly reshaping the way students acquire knowledge and develop essential competencies. Concurrently, the importance of self-regulated learning (SRL) - students’ ability to plan, monitor, and evaluate their own learning - has been widely recognized as crucial for academic success, lifelong learning, and adaptability in a dynamic labor market. Given that AI tools may significantly alter traditional learning processes, there is growing concern that reliance on these technologies might negatively impact students’ SRL skills, potentially fostering dependency and undermining critical thinking and independent problem-solving abilities.

Despite such concerns, empirical research investigating AI’s impact on students' self-regulated learning remains limited, particularly in diverse educational contexts. To address this gap, a pilot study was conducted in spring 2025, providing preliminary insights for a larger-scale investigation scheduled for fall 2025. This pilot study offers a cross-national analysis comparing higher education students enrolled in Information Technology programs in Estonia (N=48)- a digitally advanced European country - and students from Engineering programs in Serbia (N=22), a developing non-EU nation facing distinct educational challenges.

Utilizing an adapted survey instrument based on the SRL frameworks of Dawson and Guare (2010) and Strait et al. (2019), the study evaluates SRL difficulties experienced by students when using popular AI tools (e.g., ChatGPT, Grammarly, GitHub Copilot). For example, one survey item stated: "When I use AI tools, I often start assignments at the last minute because I expect the AI will help me finish faster," directly addressing procrastination behaviors linked to AI usage.

Key findings reveal significant differences between countries in terms of AI dependency and SRL behaviors. Serbian students reported greater reliance on AI for task initiation, frequently procrastinating based on expectations of AI-facilitated efficiency. In contrast, Estonian students exhibited a cautious approach, demonstrating stronger self-regulatory practices and awareness of potential pitfalls associated with AI use. Both groups acknowledged AI’s practical benefits, including improved productivity and writing skills, but also highlighted specific challenges such as distraction, loss of source tracking, and difficulties maintaining original study plans.

The results emphasize the urgent need for clear educational strategies and policies to mitigate potential negative effects on SRL skills, aligning closely with ICERI 2025's broader themes on emerging technologies, educational innovations, and skills development for future labor markets. This pilot research thus provides valuable initial insights for educators, policymakers, and researchers seeking balanced approaches to integrating AI into educational practices without compromising students’ critical lifelong learning competencies.

Keywords: Artificial Intelligence, Self-Regulated Learning, Higher Education, Educational Technology, Cross-National Comparison.