Pelindungan ciptaan yang dihasilkan oleh chat generative pre-trained transformer (Chat-GPT) (Studi gugatan class action di Amerika Serikat ditinjau berdasarkan Undang-Undang Hak Cipta di Indonesia)
Disara Wulandari, Dina Widyaputri Kariodimedjo, S.H., LL.M., Ph.D.
2026 | Tesis | S2 ILMU HUKUM JAKARTA
Penelitian ini menganalisis pelindungan hukum terhadap
ciptaan yang dihasilkan oleh Chat Generative Pre?Trained Transformer
(Chat?GPT) dengan studi kasus gugatan class action Authors Guild et
al. v. OpenAI Inc. No. 1:23?cv?08292 di Amerika Serikat yang kemudian
dikaitkan dengan kerangka Undang?Undang Nomor 28 Tahun 2014 tentang Hak Cipta
(UUHC) di Indonesia. Permasalahan yang dikaji adalah, pertama, apakah ciptaan
yang dihasilkan Chat?GPT dilindungi oleh UUHC, dan kedua, bagaimana analisis
studi gugatan class action tersebut berdasarkan UUHC.
Penelitian menggunakan metode yuridis normatif dengan
pendekatan perundang?undangan, konseptual, dan komparatif, melalui telaah
terhadap UUHC, peraturan terkait, literatur akademik, serta berkas dan
perkembangan yurisprudensi AI copyright di Amerika Serikat dan Eropa.
Hasil penelitian menunjukkan bahwa secara normatif UUHC menyediakan pelindungan yang kuat melalui pengaturan hak ekonomi dan hak moral, sementara AI tidak diakui sebagai pencipta sehingga output yang sepenuhnya dihasilkan mesin tidak otomatis menjadi objek hak cipta. Namun, belum terdapat pengaturan spesifik mengenai penggunaan karya berhak cipta sebagai training data AI serta standar kontribusi manusia dalam karya berbantuan AI. Analisis atas gugatan Authors Guild v. OpenAI dan putusan antara terkait memperlihatkan bahwa standar output?based infringement yang berkembang di AS justru mengonfirmasi bahwa pelatihan dan keluaran AI berpotensi melanggar hak cipta, dan jika diproyeksikan ke dalam UUHC posisi normatif pencipta Indonesia sebenarnya lebih kuat, tetapi masih terkendala kelemahan penegakan, forensik digital, dan kerja sama lintas negara.
This study analyzes the legal protection of authors’
works that are used works generated by the Chat Generative Pre-Trained
Transformer (Chat-GPT), using the class action lawsuit Authors Guild et al. v.
OpenAI Inc., No. 1:23-cv-08292 in the United States, which is then linked to
the framework of Law No. 28 of 2014 on Copyright (UUHC) in Indonesia. The
issues examined are, first, whether works generated by ChatGPT are protected
under the Copyright Law (UUHC), and second, how the analysis of that class action
lawsuit is conducted under the UUHC.
The research employs a normative legal methodology with
legislative, conceptual, and comparative approaches, through an examination of
the Copyright Law, related regulations, academic literature, as well as case
files and developments in AI copyright jurisprudence in the United States and
Europe.
Research findings indicate that, from a normative perspective, the Copyright Law provides strong protection through provisions on economic rights and moral rights, while AI is not recognized as a creator; thus, outputs generated entirely by machines do not automatically qualify as copyrighted works. However, there are currently no specific regulations regarding the use of copyrighted works as AI training data or standards for human contribution in AI-assisted works. An analysis of the Authors Guild v. OpenAI lawsuit and related rulings indicates that the emerging output-based infringement standard in the U.S. actually confirms that AI training and outputs have the potential to infringe copyright. When projected onto the Copyright Law, the normative position of Indonesian creators is actually stronger, but it remains constrained by weaknesses in enforcement, digital forensics, and cross-border cooperation.
Kata Kunci : Hak Cipta, Chat GPT, Artificial Intelligence, Class Action, Pelindungan Pencipta.
