Political Communication in Nepal: Artificial Intelligence, Deepfakes and Democratic Governance

Abstract

Generative tools like artificial intelligence (AI) can now create fake speeches, fake videos, and fake voices that look and sound real. This is changing the way political power is won, used, and questioned.  Using constructivist approach of review and qualitative document analysis method this paper studies how this technology, especially AI, deepfakes, is affecting political communication and democratic governance, with special attention to Nepal. In September 2025, a government ban on social media platforms triggered the Gen Z protests, the deadliest unrest in the country’s democratic history that led to the resignation of the government. Six months later, as Nepal prepared for the March 2026 House of Representatives election, the Election Commission reported widespread deepfake videos, hate speech, and AI-generated images aimed at candidates and voters. This paper reviews some global academic literature and connects it with Nepal’s National Artificial Intelligence Policy 2025, the Gen Z protests, and the 2026 election. The findings show that deepfakes damage trust in leaders, media, and elections, and this damage can be more harmful than the fake content itself. Nepal needs stronger digital literacy programs, faster fact-checking systems, clear election laws on AI content. This paper is meant to contribute by providing suggestions to policymakers, researchers, and citizens for protecting democracy in the age of artificial intelligence.