The rapid adoption of generative artificial intelligence tools in Nigerian universities has raised concerns about the erosion of student learning autonomy, yet empirical evidence on this phenomenon remains scarce, particularly regarding the curriculum and counseling responses needed to address it. This mixed‑methods triangulation study examined the extent of AI dependence and its relationship to learning autonomy among 450 undergraduate students across two universities in Imo State, Nigeria (Federal University of Technology, Owerri and Imo State University, Owerri), alongside qualitative interviews with 15 curriculum leaders, 6 counseling staff, and 20 students identified as having high AI dependence. Quantitative data were analysed using descriptive statistics and Pearson correlation, while qualitative data underwent thematic analysis. The findings revealed that AI dependence was high (88.2% used AI weekly; mean dependence score 3.72/5) and exhibited a moderate negative correlation with learning autonomy (r = –0.43, p < .001), with the strongest erosion observed in independent problem‑solving (r = –0.49). Neither university had developed coordinated curriculum or counseling responses, though the state university had piloted some AI‑resilient assessments. Students expressed guilt and a desire for guidance, indicating that institutional neglect, not student indifference, is the primary problem. The study concludes that excessive AI dependence undermines learning autonomy, and integrated curriculum‑counseling interventions are urgently needed. 

Keywords: AI dependence, learning autonomy, curriculum implications, counseling implications, higher education, generative AI, Nigerian universities, self‑regulated learning.