Adopsi Teknologi dan Tata Kelola Akal Imitasi (AI) Dalam Jurnalisme Digital: Studi Kasus Pengawasan Manusia di Ruang Redaksi IDN Times

M. YOGI E. FADILA, Dr. Wisnu Martha Adiputra, S.I.P., M.Si.

2026 | Tesis | S2 Ilmu Komunikasi

Integrasi Artificial Intelligence atau akal imitasi (AI) di ruang redaksi menghadirkan pertanyaan mendasar tentang bagaimana organisasi media mengadopsi, memaknai, dan menata kelola teknologi tersebut tanpa mengorbankan tanggung jawab jurnalistiknya. Penelitian ini mengkaji adopsi dan tata kelola AI di IDN Times untuk memahami bagaimana pengawasan manusia dirancang, diimplementasikan, dan dipertahankan di tengah adopsi AI yang masif. Fokus permasalahan terletak pada ketegangan antara efisiensi operasional yang dijanjikan AI dan akuntabilitas jurnalistik yang tidak dapat dilepaskan dari pertimbangan manusia.

Metodologi yang digunakan adalah studi kasus kualitatif tunggal dengan triangulasi tiga sumber: wawancara mendalam terhadap aktor kunci (Editor, SEO Manager, Engineer, Product Manager dan Pimpinan Redaksi) pada Februari–Maret 2026; observasi partisipatif; serta analisis dokumen internal meliputi SOP AI dan Peraturan Dewan Pers.

Temuan utama menunjukkan bahwa konfigurasi human-in-the-loop (HITL) bersifat berlapis dan disengaja sejak tahap desain. Penelitian ini juga mengidentifikasi tiga fenomena induktif baru: shadow AI, light-touch HITL, dan Generative Engine Optimization (GEO) sebagai dimensi HITL eksternal yang belum terartikulasi dalam literatur.

The integration of artificial intelligence (AI) into newsrooms raises fundamental questions about how media organizations adopt, make sense of, and govern the technology without compromising their journalistic responsibility. This study examines AI adoption and governance at IDN Times to understand how human supervision is designed, implemented, and sustained amid rapid AI adoption. The central problem concerns the tension between AI's promised operational efficiency and journalistic accountability that cannot be detached from human judgment.

A single qualitative case study was conducted with triangulation across three data sources: ten in-depth interviews with key actors — editors, product and SEO managers, engineers, and editorial leadership — in February–March 2026; two participatory observation sessions; and analysis of internal documents including the AI Standard Operating Procedure, business plans, and the Press Council regulation.

Key findings reveal that the human-in-the-loop (HITL) configuration is layered and by-design, spanning six oversight points from input curation to post-publication monitoring. AI adoption was driven by technological, organizational, and environmental pressures simultaneously, particularly Google's algorithmic changes and competitive dynamics. The study identifies three novel inductive phenomena: shadow AI, light-touch HITL, and Generative Engine Optimization (GEO) as an externally-oriented HITL dimension not yet theorized in existing literature.

Kata Kunci : Human-in-the-loop, akal imitasi, ruang redaksi, jurnalisme digital, IDN Times