The rapid rise of generative artificial intelligence (GenAI), particularly large language models (LLMs), has transformed education, prompting educators to explore ways of integrating these tools into academic workflows. While existing literature discusses their potential and challenges, little empirical research has examined students’ experiences with GenAI integration in university settings. This study presents a case from a Computer Ethics course at the Faculty of Computer Science and Engineering, where students completed their assignment implementing three writing types: fully student-written, fully AI-generated, and hybrid. A post-course, self-administered survey was conducted with 43 students, comprising 25 questions across six thematic clusters including: productivity, GenAI usage, assignment process, and ethical concerns. They combined both quantitative and qualitative components. Findings indicate that 62.79% of students were satisfied with the structured GenAI integration, citing increased efficiency and support for learning. Most used LLMs for initial idea generation and content refinement while verifying outputs with reliable sources. The analysis of open responses revealed that 41.67% had a positive experience, while concerns centered on unclear instructions, hallucinated content, and fears of plagiarism. A significant majority (72.09%) recognized ethical risks related to over-reliance on GenAI. The study highlights the need for clear instructional support, ethical frameworks, and flexible, student-centered implementation strategies. It offers practical insights for integrating GenAI in higher education and lays groundwork for future research on long-term impacts and disciplinary differences.