Analisis Ancaman Social Engineering Berbasis Agentic AI dan Strategi Mitigasi Menggunakan Pendekatan Zero Trust

Teguh Pradana (Institut Teknologi dan Bisnis Yadika Pasuruan)
Nur Amalia Hasma (Universitas Islam Kebangsaan Indonesia)
Handry Eldo (Universitas Muhammadiyah mahakarya Aceh)
Renita Selviana (Institut Teknologi dan Bisnis Yadika Pasuruan)
Ayuliana Ayuliana (Universitas Bina Nusantara)

Jurnal Ilmu Komputer dan Teknologi Informasi
Vol. 3 No. 2 (2026): September

The development of Agentic Artificial Intelligence (Agentic AI) presents new challenges in cybersecurity due to its ability to plan, adapt, and act autonomously. This study aims to analyze the threat of social engineering based on Agentic AI and to formulate mitigation strategies using the Zero Trust approach. The research method employs literature review and threat modeling to identify attack vectors, user vulnerabilities, and potential impacts. The analysis shows that Agentic AI can enhance the personalization, scale, speed, and effectiveness of social engineering attacks. This study develops a conceptual framework for mitigating social engineering attacks based on Agentic AI by applying Zero Trust principles as a layered defense approach that places identity, context, and risk at the center of security decision-making.