<![CDATA[ABSTRAK Perkembangan Artificial Intelligence (AI) berbasis Large Language Models (LLM) membawa tantangan serius terhadap integritas akademik di tingkat sekolah menengah. Penelitian ini mengeksplorasi pemahaman siswa terhadap etika digital dalam penggunaan AI untuk menyelesaikan tugas sekolah, berfokus pada anggota Paskibra SMP Negeri 2 Panongan yang memiliki karakteristik disiplin organisasi yang kuat. Menggunakan pendekatan kualitatif deskriptif studi kasus, sampel penelitian ini berjumlah 50 siswa aktif yang dipilih melalui purposive sampling. Data dikumpulkan melalui kuesioner tertutup skala Likert Modifikasi 4 poin yang telah valid () dan reliabel (). Analisis data menggunakan model reduksi, penyajian data, dan penarikan kesimpulan berbasis teori kewarganegaraan digital Mike Ribble. Hasil penelitian menunjukkan pemahaman etika digital siswa berada pada kategori Tinggi/Baik dengan rata-rata total 3,01. Berpikir kritis siswa sangat menonjol dalam memvalidasi informasi AI menggunakan buku paket sekolah. Nilai-nilai kedisiplinan dan doktrin organisasi Paskibra terbukti menjadi benteng internal siswa untuk mengutamakan daya juang mandiri sebelum mengakses AI. Namun, ditemukan ethical gaps berupa goyahnya integritas pada tugas eksakta serta kecemasan psikologis untuk melakukan transparansi penggunaan AI kepada guru. ABSTRACT The development of Artificial Intelligence (AI) based on Large Language Models (LLM) poses serious challenges to academic integrity at the secondary school level. This study explores students' understanding of digital ethics in the use of AI to complete school assignments, focusing on members of the Paskibra (Flag Troops) of Panongan 2 Public Middle School, who exhibit strong organizational discipline. Using a qualitative descriptive case study approach, the sample of this study consisted of 50 active students selected through purposive sampling. Data were collected through a closed-ended questionnaire with a 4-point Modified Likert scale that was valid (r_"count" >0.361) and reliable ("Cronbach's Alpha" = 0.863). Data analysis used a data reduction, data presentation, and conclusion drawing model based on Mike Ribble's digital citizenship theory. The results showed that students' understanding of digital ethics was in the High/Good category with a total average of 3.01. Students' critical thinking was very prominent in validating AI information using school textbooks. The discipline values and doctrines of the Paskibra organization proved to be an internal fortress for students to prioritize independent fighting spirit before accessing AI. However, ethical gaps were identified, including wavering integrity in exact science assignments and psychological anxiety about providing transparency regarding AI use to teachers.]]>
