Uji Reliansi pada AI Generatif di Arena Akademik Mahasiswa Angkatan 2024/2025 melalui Tes tanpa Pengawasan

AI generatif Pierre Bourdieu habitus arena akademik

Authors

Downloads

How to Cite

Habiburrahman, L. . (2026). Uji Reliansi pada AI Generatif di Arena Akademik Mahasiswa Angkatan 2024/2025 melalui Tes tanpa Pengawasan. Empiricism Journal, 7(2), 1117-1127. https://doi.org/10.36312/ej.v7i2.5564

Penelitian ini menganalisis fenomena reliansi mahasiswa pada kecerdasan buatan generatif (AI generatif) dalam tes soal uraian tanpa pengawasan, dengan menggunakan kerangka teori praktik Pierre Bourdieu. Studi kasus dilakukan pada 40 mahasiswa semester 2 angkatan 2024/2025 di Universitas X. Data dikumpulkan melalui analisis dokumen jawaban (840 jawaban dari 40 mahasiswa × 21 soal), wawancara mendalam terhadap 8 mahasiswa, serta studi dokumentasi kebijakan pendidikan nasional, termasuk Surat Edaran Mendikbud Nomor 4 Tahun 2020 dan kebijakan subsidi kuota internet. Hasil menunjukkan bahwa 95% jawaban memiliki tingkat kemiripan struktur dan diksi di atas 85% (diukur dengan cosine similarity). Interpretasi menggunakan konsep habitus, arena, dan modal Bourdieu mengungkap bahwa reliansi pada AI generatif bukanlah kemalasan individual, melainkan praktik rasional yang muncul dari habitus mahasiswa angkatan 2024/2025. Ini terbentuk melalui proses pembelajaran daring masa Covid-19, penghapusan Ujian Nasional, pembiasaan proyek cepat dalam Kurikulum Merdeka, legitimasi kebijakan pemerintah melalui SE Mendikbud No. 4/2020 dan subsidi kuota internet. Habitus ini kontradiktif dengan habitus di arena akademik perguruan tinggi yang masih mempertahankan logika economy reversed. Simpulan penelitian ini merekomendasikan redesain arena akademik melalui modifikasi desain tes yang memaksa produksi kognitif mandiri.

Testing Reliance on Generative AI in the Academic Context among the 2024/2025 Student Cohort through Unproctored Tests

Abstract

This study analyzes the phenomenon of student reliance on generative artificial intelligence (generative AI) in unsupervised essay-based tests, employing Pierre Bourdieu's theory of practice. A case study was conducted on 40 second-semester students of the 2024/2025 cohort at University X. Data were collected through document analysis of answer sheets (840 answers from 40 students × 21 questions), in-depth interviews with 8 students, and documentation study of national education policies, including the Minister of Education Circular Letter No. 4 of 2020 and the internet quota subsidy policy. The results show that 95% of answers exhibited structural and diction similarity above 85% (measured using cosine similarity). Interpretation using Bourdieu's concepts of habitus, field, and capital reveals that reliance on generative AI is not individual laziness, but rather a rational practice emerging from the habitus of the 2024/2025 student cohort. This habitus has been shaped by online learning during the Covid-19 pandemic, the abolition of the National Examination, habituation to fast-paced projects in the Merdeka Curriculum, and the legitimation of digital access through government policies such as Circular Letter No. 4/2020 and internet quota subsidies. This habitus stands in contradiction with the academic field of higher education, which still maintains the logic of the economy reversed. The study concludes by recommending a redesign of the academic field through modifications to test designs that enforce independent cognitive production.