Kebijakan Kecerdasan Buatan Dan Eksternalitasnya Dalam Tata Kelola Lingkungan: Tinjauan Literatur Sistematis
Jumiati Ningsih, Dede Puji Setiono, S.Pd., MDP., Ph.D.
2026 | Tesis | S2 Administrasi Publik
Perkembangan kecerdasan buatan (Artificial Intelligence/AI) yang pesat menghadirkan paradoks keberlanjutan: di satu sisi mendorong efisiensi dan inovasi, di sisi lain memunculkan eksternalitas lingkungan berupa peningkatan konsumsi energi, emisi karbon, dan limbah elektronik. Penelitian ini bertujuan untuk menganalisis eksternalitas lingkungan AI, memetakan praktik tata kelola AI dalam literatur global, serta merumuskan alternatif penguatan tata kelola AI berkelanjutan bagi Indonesia. Penelitian menggunakan pendekatan Systematic Literature Review (SLR) terhadap 43 artikel ilmiah yang memenuhi kriteria inklusi dari basis data Scopus, PubMed, dan Lens.org. yang dianalisis melalui proses coding dan sintesis tematik. Hasil penelitian menunjukkan bahwa AI memiliki peran ganda (dual role) dalam pembangunan berkelanjutan, dengan 54,3% artikel mengidentifikasi eksternalitas positif dan 45,7% eksternalitas negatif, di mana konsumsi energi menjadi titik temu utama antara keduanya. Penelitian juga mengidentifikasi dan mengklasifikasikan tiga belas instrumen kebijakan tata kelola AI global menggunakan kerangka NATO (Nodality, Authority, Treasure, Organization), yang menunjukkan dominasi instrumen Authority dan minimnya instrumen Treasure. Selain itu, ditemukan fenomena regulatory lag pada 24 artikel, 16 diantaranya mengerucut ke lima pilar dan 8 lainnya merupakan kesenjangan kebijakan secara generik. Lima pilar regulatory lag, yaitu: efek pantulan, kolonialisme digital, keterlambatan regulasi formal, ketiadaan disagregasi data energi, dan pengecualian hukum bagi otoritas publik. Dalam konteks Indonesia, keempat instrumen kebijakan yang ada (UU ITE, UU Perlindungan Data Pribadi, Surat Edaran Etika AI, dan Strategi Nasional Kecerdasan Artifisial) belum mengintegrasikan dimensi keberlanjutan lingkungan, sementara Peta Jalan Kecerdasan Artifisial Nasional dan Etika/Keamanan AI masih tertunda penetapannya. Berdasarkan temuan tersebut, penelitian ini menawarkan alternatif kebijakan melalui pendekatan policy learning, matriks kesiapan kelembagaan, dan peta jalan implementasi bertahap yang selaras dengan pencapaian SDG 7, SDG 13, dan SDG 17.
The rapid development of Artificial Intelligence (AI) presents a sustainability paradox: on one hand, it drives efficiency and innovation, while on the other, it generates environmental externalities in the form of increased energy consumption, carbon emissions, and electronic waste. This study aims to analyze the environmental externalities of AI, map AI governance practices in the global literature, and formulate alternative approaches to strengthening sustainable AI governance for Indonesia. The research employs a Systematic Literature Review (SLR) approach on 43 scientific articles that met the inclusion criteria from the Scopus, PubMed, and Lens.org databases, which were analyzed through a coding process and thematic synthesis. The results indicate that AI plays a dual role in sustainable development, with 54.3% of articles identifying positive externalities and 45.7% identifying negative externalities, with energy consumption emerging as the main point of intersection between the two. The study also identifies and classifies thirteen global AI governance policy instruments using the NATO framework (Nodality, Authority, Treasure, Organization), revealing a dominance of Authority instruments and a scarcity of Treasure instruments. Furthermore, the phenomenon of regulatory lag was identified in 24 articles, 16 of which converge into five pillars while the remaining 8 represent generic policy gaps. The five pillars of regulatory lag are: rebound effects, digital colonialism, delays in formal regulation, the absence of disaggregated energy data, and legal exemptions for public authorities. In the Indonesian context, the four existing policy instruments (the Electronic Information and Transactions Law/UU ITE, the Personal Data Protection Law/UU PDP, the AI Ethics Circular Letter, and the National Artificial Intelligence Strategy) have not yet integrated the dimension of environmental sustainability, while the National Artificial Intelligence Roadmap and AI Ethics/Security framework have yet to be formally established. Based on these findings, this study offers alternative policy directions through a policy learning approach, an institutional readiness matrix, and a phased implementation roadmap aligned with the achievement of SDG 7, SDG 13, and SDG 17.
Kata Kunci : eksternalitas lingkungan, kebijakan publik, kecerdasan buatan, regulatory lag, tata kelola AI berkelanjutan.