Elections in the Age of Digital Extremism

Elections are the heart of democracy, but voting alone does not safeguard democratic integrity. A healthy democracy depends on informed citizens who think critically and engage respectfully. Open debate, fact‑based discussion, and acceptance of diverse views form the foundations of democratic culture.

In the past, people actively sought information, compared ideas, and relied on independent media and civic participation to shape public opinion. Today, the rise of AI, algorithms, and social media has transformed how information reaches us. Digital platforms now curate content based on our preferences, creating “echo chambers” that limit exposure to diverse perspectives.

As spaces for diverse opinions shrink, outrage, accusations, and sensationalism are taking over public debate. This environment has fuelled rising extremism within democratic spaces—now one of the most serious threats facing modern democracies.

Digital Rights Nepal, Dignity Initiative, Centre for Media Research, Freedom Forum, and 20 other organizations have warned that digital disinformation could distort the pre‑election environment. They note that the three days before voting are especially sensitive and stress that physical security alone is not enough. They urge the state and stakeholders to ensure an information environment free from misinformation and manipulation.

Misinformation, manipulated content, and AI‑generated deepfakes are becoming a serious threat to voters. Such material can be created and spread within minutes, making people appear to say or do things they never did. This blurs the line between truth and fabrication, leaving ordinary voters vulnerable.

Social media algorithms often amplify sensational and divisive content. Posts that trigger strong reactions spread faster, allowing false or misleading information to go viral quickly. In today’s digital environment, the old saying “a lie travels faster than the truth” has never felt more accurate.

Another serious dimension of digital disinformation is its speed and reach. Unlike regular news or analysis, false content can reach millions within hours, quickly shaping public perception and influencing the electoral atmosphere. When voters are confused, their ability to make free and informed choices weakens—and this very judgment is what guides democracy. Algorithms now pose a real risk to that judgment.

International examples show how widespread the problem has become. In the United States, a deepfake audio‑video using President Joe Biden’s fake voice caused confusion and influenced political perceptions. In India, edited speeches and audio clips of deceased leaders circulated widely on social media, misleading voters and creating misunderstanding.

Globally, digital disinformation is expanding at an alarming pace. A 2023 LocalCircles survey in India found that 75% of respondents had encountered deepfake videos, and 27% believed election-related deepfakes to be genuine. Identity-security firm Sensity reported detecting over 250,000 deepfake items worldwide in 2022 alone.

In Europe, a McAfee study found that 10% of adults had encountered fraud involving AI‑cloned voices, with many suffering financial losses. Reporters Without Borders reported deepfake attacks against journalists in 27 countries, with women making up 74% of the targets.

In Nepal, the scale of the problem is equally alarming. Research by the Centre for Media Research–Nepal shows that 92% of social media users have encountered misinformation online. Nepal Fact Check recorded 285 potential misinformation cases between March 2020 and January 2023—57% false and 41% spread through social media. During the 2079 and 2080 elections, fact‑checkers documented widespread circulation of edited images, fake screenshots, and out‑of‑context videos. NepalFactCheck.org analysed more than 381 disinformation cases in the past year alone.

These facts show that digital disinformation in Nepal is not an isolated problem—it is widespread, organized, and a serious threat to electoral discourse. It can mislead voters, hide real information, and weaken fair democratic competition. The issue is not the technology itself, but its misuse.

The Election Commission enforced the code of conduct from Magh 4, banning misleading content, hate speech, and defamatory material—including through AI—and prioritizing a ‘Green Election’. Yet between Poush 20 and Magh 25, the e‑monitor system flagged 404 harmful items, but only 88 were removed. This gap shows how slow response, cross‑border platforms, and limited technical capacity make digital disinformation hard to control. Stronger monitoring and faster action are now essential for free and fair elections.

According to last year’s World Economic Forum report, misinformation, cyber insecurity, and social polarization rank among the decade’s biggest global risks. These threats weaken social values, erode unity and tolerance, and undermine trust in democratic institutions. Political actors have used such tactics before, prompting the Election Commission to introduce clear definitions and rules on digital disinformation in the 2079 code of conduct for the first time.

For the upcoming House of Representatives election, formal campaigning will run from Falgun 4 to midnight on Falgun 18. By law, rallies, mass meetings, corner meetings, and media‑based campaigning are permitted only within the 15‑day window before voting.

The code of conduct—developed in consultation with political parties and stakeholders—must be strictly followed. The Election Commission has adopted a zero‑tolerance policy, with penalties ranging from warnings and fines to cancelling candidacies in serious cases. Only approved advertisements may appear in print or online, and all stage‑related expenses must be submitted to the election office. Misinformation, disinformation, and hate speech on social media are strictly prohibited.

Protecting elections from digital disinformation requires a multi‑dimensional approach.

First, legal provisions must be enforced strictly. Violations of the code of conduct should be handled promptly, impartially, and transparently. Clear penalties and public disclosure help deter future violations and strengthen trust in the electoral process.

Second, coordination among key institutions must improve. The Election Commission, Nepal Police, Cyber Bureau, and Telecommunications Authority need stronger collaboration, supported by AI‑based systems that can detect and control misleading content in real time.

Third, cooperation with social media companies is essential. A clear framework is needed to ensure the rapid removal of harmful content. Only a multi‑layered strategy—combining legal enforcement, institutional coordination, and platform accountability—can effectively curb digital disinformation and protect the fairness of elections and the independence of voter decision‑making.

Long‑term solutions lie in citizen awareness. Digital literacy is no longer optional. Citizens must learn to verify sources, understand context, and engage with diverse perspectives instead of believing everything they see or hear online. Political parties must also respect voter intelligence by avoiding false promises and misinformation.

Distributing false assurances to win elections is disrespectful and deeply unprincipled. Votes won through misinformation are not democratic victories—they are moral defeats. Elections are not merely contests for seats; they are processes for earning public trust.

Democracy is built on open dialogue, critical thinking, and fact‑based public debate. Algorithms, deepfakes, and other forms of digital extremism must not be allowed to distort democratic processes or elections. Digital flows should expand dialogue, not corrupt it.

The upcoming Falgun 21 election is more than a process of choosing representatives. It is a test of the maturity and resilience of democracy in the digital age. Only when truth, conscience, and accountability remain at the centre can elections be fair, transparent, and trustworthy. Winning an election is not enough; democratic values and ethics must remain protected.