Self-Regulated Learning (SRL) is a critical skill for students' academic success and a central competency for ensuring adaptability in the labor market. While generative artificial intelligence (GenAI) holds significant potential to enhance SRL processes by serving as "metacognitive partners", the relationship between general SRL proficiency and the application of these strategies within GenAI environments remains underexplored.

This quantitative study addresses this gap by examining the disparities and correlations between learners' perceived general SRL capabilities and their perceived application of these capabilities while utilizing GenAI tools (GenAI-SRL). Data were collected via a survey involving 978 students using validated instruments to assess both general SRL and dedicated GenAI-SRL metrics.

The findings reveal a distinct and consistent gap between learners' perceptions of general SRL capabilities and SRL strategies within GenAI environments. The results indicate that the general SRL level was significantly higher than the level for GenAI-SRL, revealing a statistically significant difference with a strong effect size. Furthermore, the analysis indicates that while the frequency of GenAI use for academic purposes positively predicts the intensity of SRL strategies applied within the AI environment, the relationship between general SRL proficiency and the intentional choice to utilize the tool for learning is significantly weaker.

These findings suggest that general SRL capabilities do not automatically transfer to the GenAI environment. Despite their engagement with GenAI for academic tasks, students do not currently leverage the tool as a "metacognitive partner". Accordingly, dedicated training in GenAI-SRL is needed to assist learners in transferring their metacognitive strategies to these environments.