AI Disinformation Floods Nepal’s First Post-Protest Election

AI-generated disinformation has emerged as a significant threat to Nepal’s democratic process as the Himalayan nation heads to the polls on Thursday. These elections mark the first since deadly protests in September 2025, sparked by a brief government ban on social media platforms, which ultimately toppled the administration.

The protests that shook Nepal last year were largely driven by younger, tech-savvy citizens frustrated with chronic unemployment and rampant corruption among the country’s aging political establishment. In their aftermath, political parties across the spectrum have turned to social media to advance their agendas and court voters, particularly targeting youth and first-time voters who have registered in unprecedented numbers.

However, digital watchdogs have raised alarms about the proliferation of manipulated or entirely fabricated content circulating online. “In a country where digital literacy is low, people believe what they see,” warns Deepak Adhikari, editor of the independent fact-checking organization NepalCheck.

The challenge is particularly acute in Nepal, where approximately 80 percent of all internet traffic flows through social media platforms. According to internet analytics firm DataReportal, more than 56 percent of Nepal’s 30 million citizens are online, including 14.8 million Facebook users, 4.3 million on Instagram, and 2.2 million TikTok users.

Samik Kharel, a technology policy researcher based in Kathmandu, describes the pre-election landscape as a “digital battleground,” emphasizing that the country lacks sufficient expertise to monitor the influx of machine-generated content. “It is even hard for experts to figure out what is real and fake,” Kharel noted.

The stakes are particularly high following last year’s tumultuous events. The 2025 protests erupted after the government attempted to regulate social media, briefly banning 26 platforms including Facebook, Instagram, YouTube, and X. The resulting unrest left at least 77 people dead in just two days, saw parliament set ablaze, and ended the government of four-time prime minister KP Sharma Oli.

During that crisis, activists coordinated through the Discord app to propose an interim leader. Days later, their choice—73-year-old former chief justice Sushila Karki—was appointed to guide the country toward these elections.

The current campaign has seen sophisticated attempts at manipulation. Supporters of the ousted premier’s Marxist party have shared AI-generated images falsely depicting massive rallies, claiming attendance of over 500,000 people. These fabricated images, which TechPana’s fact-checkers determined were created using OpenAI’s ChatGPT, were subsequently shared by party leaders despite police estimates placing actual attendance at fewer than 5,000.

In another instance, an AI-generated video circulated on TikTok appeared to show Gagan Thapa, a Nepali Congress party leader, urging voters to support a rival party. TikTok has since removed the video.

The influence extends beyond Nepal’s borders. Ammaarah Nilafdeen of the US-based Center for the Study of Organized Hate noted that posts calling for the restoration of Nepal’s deposed Hindu monarchy have been circulating on social media, particularly amplified by Hindu right-wing supporters in India. These “ideological pushes” stand in contrast to domestic demands for strengthening democratic institutions.

Nepal’s Election Commission has acknowledged widespread use of hate speech and deepfake content, including videos created with readily available AI tools that falsely depict candidates insulting opponents or using vulgar language. “It is a concerning issue,” said commission information officer Suman Ghimire, who revealed that over 600 cases have been forwarded to authorities, with approximately 150 handled by police.

In one case, authorities detained pro-royalist supporter Durga Prasai for social media posts allegedly intended to intimidate potential voters. While the Election Commission can impose fines or disqualify candidates, experts believe the sheer volume of online disinformation far outpaces any effective regulatory response.

“Candidates and people close to political parties not only compete to win, but also compete to spread misinformation,” observed Basanta Basnet, editor-in-chief of the news website Onlinekhabar, which collaborates with Nepal FactCheck to verify posts. The organization has warned that misinformation could lead citizens to make ill-informed decisions, potentially undermining the very “foundation of democracy.”