Misinformation During Nepal’s 2026 Elections

Executive Summary

Nepal’s 2026 House of Representatives elections took place in a volatile political environment following the 2025 Gen Z movement, the collapse of the previous government and declining public trust in political and media institutions.

Against this backdrop, the study examined the nature, patterns and dynamics of political misinformation in the month leading up to the elections. Drawing on a dataset of 102 fact-checked misinformation cases and 1,020 instances of potential misinformation recorded across all seven provinces, the study provides the first systematic analysis of election-related misinformation in Nepal. As misinformation is a key factor in undermining peace by fuelling fear, prejudice, polarisation and social tensions, particularly during politically sensitive periods such as elections, the study is also deemed relevant to peacebuilding professionals working in digital peacebuilding.

The findings show that misinformation was widespread and overwhelmingly disseminated through social media platforms, with Facebook being the most prominent misinformation vector. Most of the misinformation pieces were not entirely fabricated but involved the manipulation of authentic content through misleading captions, selective editing, recycled images and videos and other forms of context distortion. Election misinformation was strongly leader-centric, focusing on prominent political figures rather than policy issues and relied heavily on visual content designed to provoke emotional reactions and maximise engagement. Although AI-generated content, including deepfakes and synthetic media, was present, it remained less common than low-cost forms of manipulation such as edited images, videos and so-called “cheapfakes.”

This study also finds that misinformation targeting women politicians differs from that targeting male political leaders. Gendered misinformation frequently targets women’s private lives, family relationships, personal character, appearance and social status. Gendered misinformation was particularly evident on social media platforms, where anonymous and pseudonymous accounts circulated misleading narratives about women politicians, endangering women’s political participation.

The study further found that misinformation frequently exploited narratives related to political leaders, electoral violence, foreign interference and electoral fraud. Such narratives appeared particularly during the final stages of the campaign and were often designed to undermine trust in democratic institutions and electoral processes. Monitoring data indicates that misinformation activity increased significantly as election day approached, suggesting that periods of political competition create favourable conditions for the spread of misleading content.

The findings suggest that Nepal’s misinformation ecosystem is a result of political polarisation, declining trust in media, the growing influence of social media platforms and content creators, weak media and information literacy and limited fact-checking capacity. While existing interventions such as fact-checking, monitoring and public awareness campaigns contributed to addressing misinformation, they remain fragmented and insufficient. The study recommends a coordinated multi-stakeholder approach involving state institutions, civil society, media organisations and digital platforms to strengthen information integrity, enhance media and information literacy campaigns, improve fact-checking capacities and safeguard electoral integrity in Nepal.

1. Introduction

Nepal held the general elections on March 5, 2026. This election followed the dissolution of the House of Representatives by President Ram Chandra Paudel and the formation of a caretaker interim government under former Chief Justice Sushila Karki. The need for a general election arose from the political fallout of the Gen-Z movement, which took place in September 2025. The Gen Z movement was a social-media-driven movement fuelled by years of youth frustration with governance and political instability (Acharya, 2025). The government led by former Prime Minister K. P. Sharma Oli enforced a ban on Facebook and 25 other major social media platforms on 4 September 2025 after the platforms failed to register with the Government of Nepal under the country’s social media regulatory framework, despite repeated notices. The ban had severe implications for communication, livelihoods and freedom of expression. The digital-native Nepali youths, while discussing the social media ban, called for peaceful demonstrations. These demonstrations took place on September 8 and quickly escalated into nationwide protests with broader demands for government accountability and an end to corruption (Koirala, 2025). The unrest culminated in the death of at least 75 people, nationwide acts of vandalism targeting government buildings and private properties and the resignation of Prime Minister KP Sharma Oli. Given this volatile political situation, the 2026 general elections set for March 5 were vulnerable to the influence of misinformation, as “Nepal’s information ecosystem was facing a critical juncture characterized by an alarming rise in misinformation and information manipulation” and “the digital public sphere increasingly being used as a site for political polarization and the spread of false narratives” (Dahal, 2025).

Globally, the rise of the internet has altered the role of the public, shifting from passive consumers to active producers of political content. This active participation presents considerable challenges to democratic processes, particularly through social media platforms, which frequently foster echo chambers and reinforce homogeneous viewpoints (Bhandari, 2024). In Nepal, specifically, the emergence of social media platforms has been linked to shifts in political discourse, with platforms like TikTok, Facebook, YouTube and X (formerly Twitter) being used as key platforms for the dissemination of both authentic and fabricated information. The 2025 Gen Z protests highlighted the dual nature of these platforms. They enabled quick mobilization for democratic engagement while also showing their vulnerability to misinformation, which can escalate societal tensions and impact political stability (Timilsina, 2026). This, along with declining public trust in traditional media, has created fertile ground for the proliferation of manipulated narratives, a concern amplified by the potential for AI-driven misinformation to disrupt democratic processes and influence electoral outcomes (Acharya & Banu, 2025). The prevalence of misinformation during critical periods, such as elections, disasters and public health crises, has been linked to organized cyber troops established by political parties, as well as the strategic dissemination of false narratives by foreign governments (Dahal, 2025). The global landscape already recognizes misinformation as a pressing challenge for democracies and while misinformation is not new, its prevalence and potential for harm have been significantly amplified by the disruptive influence of social media platforms on communication systems (Acharya & Banu, 2025).

In Nepal, this transformation is marked by 15% annual growth in internet penetration, with Facebook reaching an estimated 11.85 million users (Rupakheti, 2025). Specifically, in Nepal, an estimated 13.5 million individuals actively engage with social media, constituting 49.6% of the entire population, thereby establishing these platforms as primary conduits for both political expression and the widespread dissemination of diverse content, including misinformation (Bhandari, 2024; Kharel, 2024). With the increasing growth of social media platforms, the mainstream media has faced an existential threat in Nepal (Sharma, 2025). Given their broad and diverse user base, social media platforms play a key role in political mobilisation and shaping public discourse.

Against this backdrop, this study examines the nature, patterns and dynamics of misinformation circulating in the lead-up to Nepal’s 2026 general elections. Digital spaces have emerged as “a fertile ground for fake news campaigns and online hate speech,” with political contexts becoming especially conducive to misinformation, often amplified by social media platforms (Vasist et al., 2024). The 2026 general election in Nepal occurred at a time when public trust in state institutions was fragile, established political institutions and actors were weakened and social media platforms remained a key platform for the nation’s political discourse. Ongoing monitoring of misinformation trends is essential in Nepal’s evolving information ecosystem, particularly in the lead-up to high-stakes events such as elections (BBC Media Action, 2025). As such, understanding how misinformation played out in this environment is important to assess Nepal’s digital public sphere.

While misinformation has been widely studied in the context of elections in established democracies, empirical research on election-related misinformation in emerging digital democracies is limited. The unique circumstances surrounding the 2026 general elections provide a rare opportunity to examine how misinformation impacts elections during a period of political crisis and change. This study documents the types of misinformation, actors and narratives, and seeks to contribute to policy discussions on safeguarding electoral integrity in digitally mediated political environments, thereby adding to the broader debates on information integrity and democratic resilience in the Global Majority.

The study is organised as follows: the introductory chapter sets out the political background, the 2025 Gen Z movement, objectives and research questions, the significance of the research, biases and limitations and the key terminologies. This is followed by the literature review, which situates the study within the context of existing research on misinformation, the media landscape, information consumption and the interventions deployed to counter misinformation in Nepal. The findings chapter presents the analysis in three parts: confirmed misinformation drawn from the fact-checked dataset, potential misinformation drawn from the monitoring dataset and a comparison with the 2022 general elections. The report closes with conclusions and a set of recommendations for state institutions, civil society organisations and the media.

1.1 Political Background

Nepal’s political history is turbulent. Modern-day Nepal was established in 1768 AD by King Prithvi Narayan Shah (1723–1775), who unified numerous small states, including the Kathmandu Valley’s three city-states ruled by the Malla kings. A power struggle within the palace led to the rise of the Rana regime in 1846, which was an absolute hereditary rule of Rana prime ministers, while the monarchy remained a figurehead. An armed struggle led by the Nepali Congress, amidst a difficult political situation in which King Tribhuvan Shah fled to India seeking refuge, ended the Rana rule and introduced democracy with a constitutional monarchy in 1951, following a tripartite agreement among King Tribhuvan Shah, Rana Prime Minister Mohan Shamsher and the Nepali Congress in New Delhi, India.

Since then, Nepal’s political system has remained in a transitional phase, with major political movements creating ripples periodically. In 1962, King Mahendra introduced a party-less Panchayat system, removing multiparty democracy, following the 1960 royal takeover. The monarchical system faced a major challenge in 1980 when King Birendra Shah was forced by student protests to call a national referendum to choose between a reformed party-less system or multiparty democracy. The narrow victory in the referendum was only able to sustain the Panchayat system for another decade, as in 1990, a people’s movement or Jana Andolan I led by the Nepali Congress and the United Left Front, a coalition of communist parties, forced the King to restore multiparty democracy under a constitutional monarchy.

The Communist Party of Nepal (Maoist), in 1996, began an insurgent movement, known as the Civil War, that lasted a decade and resulted in the killing of over 17,000 people. A Royal Massacre in 2001 at the palace, which saw the reigning king, queen and crown prince killed, weakened the monarchy and led to the enthronement of King Gyanendra Shah. King Gyanendra dismissed the government and assumed direct rule in 2005, resulting in a people’s movement called Jana Andolan II, which was also joined by the Maoist party in 2006, forcing King Gyanendra to reinstate parliament. This was followed by the Maoists joining mainstream politics after the Comprehensive Peace Accord. In 2008, the Constituent Assembly officially abolished the monarchy and established Nepal as a Federal Democratic Republic. A new constitution was adopted in 2015, establishing Nepal as a secular Federal Republic.

Since the 1990s, Nepal’s political landscape has been dominated by three political forces, the Nepali Congress, the United Marxist Leninist and either the King or the Maoists. The political system has been characterized by coalitions and political instability, with power struggles, frequent government changes and uncertainty. This cycle ended in September 2025 when the Gen Z movement led to the fall of the government and the formation of a new interim administration mandated to conduct elections on March 5, 2026.

1.2 The 2025 Gen Z Movement in Nepal

The youth-led protests in Nepal in September 2025, referred to as the Gen Z movement, is an example of digitally mediated political mobilization in the Global Majority. Similar movements had been seen in the last two years in Bangladesh, Sri Lanka, Indonesia, Madagascar, Morocco, the Philippines, Kenya and Timor-Leste. The primary characteristics of these movements are youths’ dissatisfaction with the economy, stagnant political situation and corruption, as well as their mobilization via digital platforms. The protests were fuelled by the introduction of new laws or regulations and were aimed to put pressure on the government for a systemic change. Nepal’s Gen Z movement demonstrated the significant potential of young digitally literate citizens to exert pressure on established political leadership (Timilsina, 2026). The demonstrations were organized and led by young citizens born after 2000 AD who had long-standing grievances with their country’s governance.

Nepal’s protests were triggered by the government’s decision on 4 September 2025 to ban 26 major social media platforms as they did not comply with the government’s demands to register locally. This regulatory registration obligation was taken by digital rights activists as a restrictive step and ill-guided, intended to exert control over social media platforms to punish criticism. The government issued rules requiring social media platforms to register locally and remove government-flagged content within a specified timeframe or face penalties. For many Nepalis, this decision meant not only a regulatory enforcement but a restriction on communication, entrepreneurship and freedom of expression. Social media platforms had become central for communication, small business promotion, content creation and maintaining social ties with migrant family members abroad, making the ban highly disruptive to everyday social and economic life (Acharya, 2025).

While deeper issues existed, the government’s decision to ban or restrict social media served as the final catalyst that brought large parts of Nepali youth onto the streets to defend their digital freedom (Dahal, 2026). Political instability, corruption and their exclusion from political decision-making had created widespread dissatisfaction among younger citizens. The movement has revealed the depth of problems, including failures to provide political leadership, the decay of traditional political authorities and neglect to comply with a structural reform agenda (Sapkota, 2026). For decades, Nepal’s political landscape has been dominated by a handful of high-ranking politicians who take turns in power through a series of coalitions, thereby leaving no room for a generational change in leadership. This persistence of elite political control, combined with bureaucratic inefficiency and corruption in public institutions, has contributed to declining public trust in the state and heightened frustration among younger generations (Acharya, 2025).

Against this background, youths sharing their frustrations over online discussions across multiple digital platforms organized the demonstrations. Initial conversations about the protests took place on social media sites and discussion platforms such as Reddit before moving to Discord, where participants coordinated plans for demonstrations and delegated responsibilities. This digitally mediated organization of protests reflects what Bennett & Segerberg (2013) describe as connective action, in which political mobilization occurs through personalized digital networks rather than traditional hierarchical organizations. On 8 September 2025, youth groups gathered for a peaceful anti-corruption protest at Maitighar in central Kathmandu. However, the situation quickly escalated when demonstrators stormed the parliament compound and security forces responded with heavy-handed measures including tear gas, rubber bullets and live ammunition (Acharya, 2025).

After the first day of violence instigated by the state, which resulted in the death of 19 young demonstrators, including students, the movement rapidly spread to various regions of the country the following day (Sapkota, 2026). Demonstrators attacked state institutions and properties associated with political elites, while the government reacted by imposing curfews and deploying the military to restore order. Amid escalating unrest, the Prime Minister resigned and an interim citizens’ government was formed (Acharya, 2025).

A central aspect of the Gen Z protests was the role of digital information flows, including the spread of misinformation during the crisis. The rapidly evolving situation created an information vacuum in which rumours, manipulated information and misleading narratives circulated widely. These included misattributed videos from other countries presented as events in Nepal, exaggerated casualty figures, false claims of military coups and misleading narratives linking the protests to religious or nationalist agendas (Acharya, 2025).

The prevalence of misinformation during the Gen Z movement highlighted a broader challenge with digital activism. Such environments are often characterized by information overload and emotional amplification, which can facilitate the rapid spread of misleading content (Boyd, 2014). The speed and scale of online information stressed the limited ability of citizens to distinguish between verified information and rumours during moments of crisis. With mainstream media in Nepal losing ground and reach due to, “traditional news reporting methods, delayed dissemination, formality, paid news practices, and focusing on the interests of elites […] actually the major reasons why mainstream media are losing ground and scope in Nepal”, (Sharma, 2025) the stage was set for the spread of misinformation on an unprecedented scale via social media platforms. The diminishing public trust in mainstream media also contributed to the spread and impact of misinformation.

1.3 Objectives of the Study

An effective response to misinformation requires multidisciplinary approaches. Digital and media literacy initiatives, training for journalists, fact-checking systems and policy interventions should ideally work in tandem instead of being implemented in silos. However, any such intervention must be based on a systematic analysis and understanding of the misinformation patterns and narratives that dominate the information landscape of the respective national context. The study hence focuses on analysing prevailing narratives, frames and actors of political misinformation to assess the state of misinformation in the Nepalese digital sphere.

Hereby, the patterns, framings and narratives of misinformation disseminated on the social media platforms YouTube, Facebook, X (formerly Twitter) and TikTok in the run-up to the Nepalese parliamentary elections on 5 March 2026 have been analysed. The timeframe for the social media content analysis, which was carried out one month prior to the elections, was deliberately chosen since, before elections, a heightened level of societal polarization can generally be observed. This trend is usually accompanied by a sharp increase in the spread of misinformation on social media platforms. YouTube, X, Facebook and TikTok were selected as platforms for the content analysis, given they have the largest user base in Nepal and thus the greatest reach. By focusing on the four platforms with the largest user base, we aimed to ensure that the results reflect the narratives, framings and patterns of political misinformation that impact a significant portion of the Nepalese population, rather than just fringe segments of society. This ensures the study’s broader applicability to the practical contexts as well as its significance.

Furthermore, misinformation is rightfully described as a global phenomenon, yet little attention is paid to the fact that it is equally a context-bound phenomenon. As stated in a study on misinformation conducted by Zhang and Ghorbani, this social context is pivotal to the success of misinformation’s spread and its influence on the wider public. Hereby, social context should be understood as encompassing all aspects of the social, cultural, political, economic, and environmental landscape, including key events and trends within a particular society. Hence, the general objective of the following study is also to capture and analyse this social context of misinformation dissemination in Nepal.

Additionally, the study shall address the following specific objectives:

• Identify and document dominant political misinformation narratives circulating on social media around Nepal’s 2026 elections.
• Examine how these narratives are adapted for different audiences and target groups.
• Share findings into patterns of political misinformation, including emotional tone, timing and coordination.
• Generate evidence-based recommendations for policy and operational interventions, with particular emphasis on civil society engagement and media literacy.

1.4 Research Questions

The study will provide answers to the following interconnected research questions:

• What major political misinformation narratives circulated in the digital space during Nepal’s 2026 elections?
• What patterns and actors are observable across the selected social media platforms?
• Who are the actors responsible for amplification of political misinformation?

1.5 Significance of Research

The study provides baseline information on the current state, actors and narratives of political misinformation on social media platforms in Nepal. It provides insight into the genesis, framing, actors and dissemination of misinformation within Nepal’s social media information ecosystems. Such a contextual understanding of misinformation patterns supports policymakers, social media consumers and civil society organisations alike to develop fact-based approaches, interventions or policies to detect, counter and prevent the spread of misinformation. Hereby, it also ties in with broader efforts to strengthen digital and media literacy, platform governance and the promotion of reliable information environments for social media users.

As a country with a history of armed insurrections and political unrest – most recently the Gen Z protests in September 2025 against political corruption – misinformation spread on social media can have tremendous offline ramifications, leading to real-life violence. Hence, combating misinformation in Nepal’s post-conflict context goes beyond digital information integrity, directly contributing to broader peacebuilding and conflict prevention initiatives. Indeed, within the Nepalese context, combating misinformation should be regarded as intertwined with peace-building objectives, as it has become a pivotal factor towards establishing a peaceful society.

Written by media literacy practitioners who have been working in this field within the Nepalese context for several decades, the study also offers a realistic, nuanced and practice-oriented assessment of effective measures to combat the spread of misinformation. It is the first study of its kind to analyse narratives, patterns and actors of political misinformation in the run-up to elections in Nepal. Furthermore, the study provides a basis for further research in this field in other multi-ethnic, multilingual and multi-religious post-conflict societies in the Global Majority.

1.6 Biases and Limitations of the Study

The findings of this study should be interpreted in light of several limitations relating to data, methods, scope and potential biases. The most important limitation is that the study does not examine the impact of misinformation on voter behaviour or voters’ electoral choices. The used datasets are also selective. The fact-checked misinformation dataset represents only the subset of misinformation that fact-checking organisations chose to verify and publish, while the monitoring dataset consists of content flagged as potentially false or misleading. Although potential misinformation was monitored across all seven of Nepal’s provinces, the sample is not nationally representative, as it comprises only content encountered and recorded by the monitors on their timelines.

As a result, the findings are subject to selection bias and may not represent the complete misinformation ecosystem. Content shared through private or encrypted channels such as Facebook Messenger, WhatsApp, Viber, and Telegram remained outside the study’s scope. The analysis is further constrained by its focus on a limited number of platforms during a one-month campaign period, capturing election-related activity without comparison to non-election periods. While the study identifies plausible links between misinformation, declining trust, polarisation and risks to electoral integrity, these relationships should be understood as associational rather than causal. Comparisons with the 2022 elections should also be interpreted cautiously due to differences in fact-checking capacity and data collection methods. The findings are supplemented and validated by the perspectives of focus group discussion media experts, whose expertise, whilst providing valuable insights, remains subjective.

Finally, the study was conducted by researchers who are also practitioners engaged in public awareness and information integrity initiatives. Although every effort was made to maintain analytical objectivity, the possibility that the researchers’ professional experiences may have influenced certain interpretations cannot be entirely excluded.

1.7 Terminology

• AI-generated content (or synthetic media): Images, audio or video created completely or partly by generative artificial intelligence tools.
• Cheapfake: Misleading content produced through low-cost, low-skill manipulation of real material such as selective editing, miscaptioning, recycling old footage or presenting unrelated visuals as evidence.
• Connective action: Political mobilisation that occurs through digital networks and the sharing of content rather than through traditional organisations.
• Content creator / influencer: An individual or account with significant social media following that produces and circulates content, including political content, often without following professional journalistic standards.
• Context manipulation: Misrepresenting real images, videos, speeches or events to alter their meaning by changing the context using misleading captions, false dates, clipped quotations or unrelated pairing.
• Coordinated inauthentic behaviour: The use of fake, impersonating or automated accounts to artificially amplify content on social media platforms.
• Cyber troops: Teams established by political parties or state actors to shape online discourse, sometimes promoting misleading content.
• Deepfake: Synthetic audio, image or video content generated or significantly altered using artificial intelligence.
• Digital public sphere: The online spaces, including the social media platforms, in which public political discussions happen.
• Echo chamber: An online environment in which users are exposed to information and opinions that reinforce their existing views.
• Fact-checking: The systematic verification of public claims and viral content against available evidence.
• False / fabricated content: A fact-check classification applied to content that is entirely invented or synthetically produced and has no connection to any real event, statement or image.
• Gendered misinformation: Misinformation that targets individuals on the basis of gender, such as attacking women in politics through sexualised content, attacks on private life and character assassination.
• Global Majority: A term used to refer to the countries and populations that constitute the numerical majority of the world.
• Impersonation / fake accounts: Accounts created using another person’s name, image or identity (impersonation) or wholly fictitious identities (fake accounts).
• Information disorder: A term for the information ecosystem where misinformation is abundant.
• Information ecosystem: The overall environment in which information is produced, distributed and consumed within a society, including mainstream media, social media platforms, content creators, audiences and the technological and regulatory systems.
• Information integrity: The reliability, accuracy and trustworthiness of information.
• Media and Information Literacy (MIL): The competencies that enable citizens to access, critically evaluate, verify and contextualise information.
• Misinformation: An umbrella term for false or misleading content, irrespective of the intent behind its creation or spread.
• Misleading content: A fact-check classification applied to altered real material to convey a false impression.
• Peacebuilding / conflict transformation: Processes aimed at preventing violence and building the social and institutional trust necessary for durable peace.
• Political misinformation: False or misleading information that is either political in nature and/or relevant in the context of political events, and/or shared for political benefit.
• Social-first media channels: Outlets and pages that originate and operate primarily on social media platforms, some of which present themselves as news media despite lacking editorial standards.
• Unverified content: Content that fact-checkers could neither confirm nor disprove with the evidence and tools available.

2. (Mis)information Landscape of Nepal

2.1 Misinformation in Nepal

The misinformation landscape in Nepal has become a significant challenge to the country’s democratic discourse and social stability. Misinformation poses a significant threat to Nepal’s society and democracy, as the country’s deep ethnic, linguistic and religious diversity, combined with a history of exclusionary nation-building, creates a fragmented environment vulnerable to identity-based manipulation. As Pathak & KC (2025) argue, long-standing political instability as well as the geopolitical situation make Nepal a fertile ground for misinformation campaigns. This digital expansion contributed to the proliferation of falsehoods, particularly on popular social media platforms such as TikTok and YouTube, which are increasingly replacing traditional media as primary information sources (Dahal, 2025).

This situation became evident during the Gen Z movement. Misinformation was a significant challenge as it intensified panic, distorted real events and amplified violence in an extremely volatile environment. Various newspaper reports mentioned a flood of misinformation circulating rapidly during the protests, exploiting public uncertainty and the demand for real-time information. This resulted in increased confusion and fear among citizens (Roy, 2025). False claims, including fabricated reports of attacks on religious sites, deaths of public figures, and misrepresentation of protester demands, spread widely across social media and were further amplified through cross-border information flows, shaping public perception beyond Nepal (Global Issues, 2025). Media reports citing research findings suggested that a significant share of online discourse, estimated at around one-third, was driven by fake or coordinated accounts, artificially amplifying specific narratives and potentially influencing the trajectory of the protests (Online Khabar, 2025). Although misinformation was spread abundantly previously during political events in Nepal, its speed and volume during the Gen Z protests has been unprecedented. This contributed to polarization and made it difficult for citizens and authorities alike to distinguish credible information from falsehoods (Ray, 2025).

This relationship between misinformation and real-world violent consequences is not new in Nepal, as there are historical examples. Nepal’s experience with misinformation-driven violence predates the rise of social media. The Hrithik Roshan Scandal in December 2000 being the most cited example. A fabricated claim that Bollywood actor Hrithik Roshan, in a television interview, said he hated Nepal and its people spread through word of mouth, FM radio and print media, which escalated into nationwide protests. As Bhattarai (2023) recounts, demonstrations that began locally in Chitwan spread rapidly across the country, resulting in at least four deaths, widespread vandalism and disruption of daily life. A government inquiry later found no evidence that such a statement had ever been made. This incident shows how misinformation activates existing grievances, such as sensitivities around Nepal–India relations, to incite violent reactions and demonstrates that the relationship between misinformation and unrest is rooted in social vulnerabilities.

A second case from 2009 shows how misinformation can incite decentralised incidents of violence across the country. Nepali Times (2009) reported that in June and July 2009, 18 individuals were killed in various parts of Nepal following the circulation of misinformation that unidentified individuals were attempting to abduct children. Investigations later confirmed that the victims were innocent. The violence, which began in the Tarai and spread to the Kathmandu Valley, showed the speed with which unverified claims can trigger mob action. These incidents indicate that misinformation exploits social distrust, weak institutional confidence and information vacuums.

Digital platforms have intensified the link between misinformation and communal tensions. For example, in October 2023, inflammatory content circulating on digital platforms incited communal clashes in Nepalgunj, prompting the imposition of an indefinite curfew and high-level security arrangements to stop the violence from spreading (The Kathmandu Post, 2023). A similar incident happened in January 2026, when a TikTok video allegedly containing offensive religious remarks spread rapidly, inciting protests and violence in Dhanusha and Parsa districts, including unrest in the city of Birgunj (Online Khabar, 2026).

These incidents show that online content can trigger violent offline consequences, particularly in socially sensitive regions, where complex religious and geopolitical tensions are present. They suggest that the spread of misinformation in the country poses a big risk to peace and harmony, as fragile socio-political contexts are present.

The misinformation landscape is characterized by a diverse range of threat actors, including political parties, extremist groups and even government or foreign entities (Dahal, 2025). During election cycles, the spread of political propaganda and conspiracy theories through unregistered news portals significantly intensifies (Prajapati et al., 2025). Furthermore, technological advancements have introduced new risks; for instance, the emergence of AI-generated deepfakes has been identified as a modern aggravating factor in Nepal’s misinformation ecosystem (Acharya, 2025). Evidence from multiple CMR-Nepal surveys from 2017 to 2023 indicates the widespread prevalence of misinformation in Nepal:

“Surveys consistently find high exposure and concern among internet users: in 2019, 95.5% of heavy internet users reported encountering misinformation within a week and 86.5% expressed concern; in 2021, over 91% reported recent exposure and 86.8% remained concerned; and even among non-heavy users in 2023, 68% had encountered misinformation recently. Across these studies, an overwhelming majority, ranging from around 96% to nearly 98%, believe misinformation is already or will become a serious problem for society and politics, indicating that misinformation is not only omnipresent but also recognized as a major challenge.”

— Pathak & KC, 2025

Several factors contribute to this situation. High levels of political instability and low public trust in traditional media have weakened the country’s information integrity (Acharya, 2025). Traditional newsrooms are facing a financial crisis and hence often succumb to the pressure of social media influencers and clickbait journalism that prioritizes viral content over factual accuracy (Dahal, 2025).

The psychological consequences of misinformation are gaining global recognition. Some researchers argue that citizens may be suffering from a so-called “information disorder syndrome” due to the constant exposure to misleading content (Kharel, 2020). The situation is complicated by the lack of accountability from global social media platforms, which often prioritize larger markets and lack robust, country-specific regulatory mechanisms for Global Majority countries like Nepal (Acharya, 2025).

Evidence from Nepal shows that misinformation is often gendered (Purple Foundation, 2022). A study of the Purple Foundation found that anonymous rumours, hate speech and misinformation are systematically directed at women politicians, particularly during elections, and such narratives are a politically motivated strategy to exclude women from political participation. The research further highlights that 63% of respondents, who were women in politics, experienced misinformation and hate speech during elections.

A study conducted by Panos South Asia (2022) shows that social media is the primary arena where gendered misinformation and abuse against women politicians unfolds in Nepal. The 2022 study documented 243 cases of misogynistic content, with an overwhelming majority appearing on platforms such as Facebook, Twitter, and TikTok. The research highlighted that negative comments, hate speech, trolling and verbal abuse are widespread on social media platforms and mostly carried out by anonymous accounts. Gendered misinformation intensifies during election periods, where insults and hate speech are often intertwined with misinformation and are being used to undermine women politicians’ credibility and participation (Panos South Asia, 2022). This evidence shows that gendered misinformation is not only present in Nepal, but indeed does harm individuals and democratic participation.

2.2 Media Landscape and Information Consumption

Nepal’s contemporary media landscape stands at a crucial juncture, coming under pressure due to the digital transition of the information ecosystem, economic hardships and a loss of public trust in the face of ever-increasing misinformation (Deutsche Welle, 2024). Even though mainstream media in Nepal has been historically partisan to the level that most Nepali citizens perceive journalists as political instruments rather than objective voices, public trust in the media system is continuing to deteriorate (Pathak & KC, 2025).

Over the last decade, mainstream media has gradually adopted digital-first strategies by moving to internet-based platforms to disseminate their content. Thereby new online media outlets and social-first media channels have emerged. This is because the internet has penetrated Nepali society at an incredibly high speed due to the spread of mobile technologies and cheaper smartphones. According to the Nepal Telecommunications Authority, the internet penetration in late 2025 stood at 144 percent of the population, with 88.8 percent using the internet through mobile broadband services.

Nepal’s current media industry comprises more than 10,000 registered outlets across the country, more than half of which are concentrated in the Kathmandu Valley. Additionally, there are nearly 10,000 journalists who are union members.

In recent years, the growth of online media has been exponential. Between 2017, when the government Department of Information listed less than 1,000 online outlets, to 2025, the number of online media outlets increased fivefold. At the same time, print newspaper circulation figures are falling. Out of 5,000 registered newspapers, only about one-third are regularly published and the remaining are either irregular or digital-only. According to the annual report by the Press Council Nepal, in 2024/25, the number of new registrations for online media stood at 500, while print registrations remained stagnant. Many radio and television stations are operating under severe financial constraints, focusing more on low-cost content such as music and feeder content rather than exclusive journalistic pieces.

In contrast, the social-first media channels, especially those on Facebook, Instagram, TikTok and YouTube, have become influential information dissemination channels to many people using mobile phones to access information. Many social media influencers have institutionalized their channels as media entities, although they lack journalistic standards and rigor in content production. At the same time, it is their channels that are becoming the main source of information for many Nepali citizens who were previously not direct consumers of mainstream media. Many of these content creators, however, are not following professional journalistic ethics, providing ample avenues for the flow of misinformation (Anupam, 2025). As social media increasingly becomes the primary source of information, information disorder is growing (The Asia Foundation, 2024). The prevalence and influence of misinformation have also undermined public trust in information received through mainstream media.

The economic challenges faced by the media sector have been widely documented, especially the decline in traditional revenue from advertising and print sales, leading to reduced newsroom capacity and layoffs. The traditional revenue model based on advertisements from the government and private sector, such as print sales, has almost collapsed. Now advertisers are shifting funds to digital and social platforms, reducing income for newspapers, radios and television channels, as well as forcing some outlets to stop print editions and cut editorial staff. The economic pressure has also contributed to increasing working hours and pressure to produce more content, as many journalists have faced pay cuts, unemployment, or reduced operational support, weakening the professional environment and quality of journalistic output. As a result, the quality of journalistic content has become erratic, with fewer resources available for in-depth reporting and editorial standards under pressure. The Press Council Nepal (PCN) received 341 formal complaints last fiscal year regarding character assassination, lack of verification and yellow journalism.

The Constitution of Nepal of 2015 guarantees full press freedoms in its preamble. Relevant fundamental press rights are enshrined in Article 17 (right to freedom of expression and opinion), Article 19 (right to communication) and Article 27 (right to information). However, the actual policy environment in recent times has been increasingly restrictive. The government has attempted to introduce policies that are restrictive to both freedom of expression and press freedom. The Media Council Bill sought to replace the self-regulatory nature of the Press Council with a government-controlled body, whereas the Social Media (Usage and Regulation) Bill, now withdrawn, sought to punish citizens for their critical commentary on social media platforms.

Nepal’s state of press freedom in 2025 was described as finding itself in a “distressing and deepening crisis” marked by a decline in global rankings. Reporters Without Borders (RSF) downgraded Nepal to 90th place in the 2025 World Press Freedom Index, a significant fall from its 74th position the previous year, citing a “press freedom crisis” fuelled by political instability and economic fragility. Similarly, Freedom House, in its 2025 report, maintained Nepal’s status as “partly free”.

2.3 Interventions to Combat Misinformation and Challenges

Nepal’s response to misinformation in electoral contexts has evolved gradually since 2017. Early efforts were exploratory and reactive, but over the years, they have expanded into monitoring, fact-checking, media literacy initiatives and regulatory attempts. However, the interventions remain fragmented and insufficient to match the volume and speed of misinformation.

Ahead of the 2017 general elections, civil society organizations began studying misinformation and documenting its patterns. CMR-Nepal published a flyer targeted to journalists on misinformation and monitored social and mainstream media for potential misinformation. During the 2017 general elections, CMR-Nepal found that misinformation was becoming a political tool to insult, accuse and demean opponents, as well as to spread false information to impact the popularity of favourable candidates (Pathak & KC, 2025).

In 2022, Nepal held elections at the local, provincial and federal levels. The electoral cycle witnessed diverse media use, including misuse, the spread of misinformation and occurrences of hate speech (The Asia Foundation, 2024). Interventions by then had also expanded. The Election Commission of Nepal introduced guidelines and monitoring mechanisms targeting online content (NDI et al., 2024). Civil society organizations monitored misinformation, conducted public awareness campaigns and research studies. Fact-checking initiatives such as NepalFactCheck.org and NepalCheck.org stepped up operations to verify viral claims. However, it should be noted that fact-checking is a reactive method for addressing misinformation. Research suggests that misinformation corrections lose effectiveness rapidly within a short timeframe because “corrective messages quickly fade from memory,” (Brashier et al., 2021) something like “after a 1-week delay was worryingly low” while misinformation itself has “continued influence effect.” (Ecker et al., 2020) Despite policy efforts from the Election Commission of Nepal and the monitoring and public awareness initiatives from civil society and mainstream media, misinformation was rampant. There were fake screenshots of established news media outlets, old photos, morphed videos and out-of-context contents used to degrade opponent candidates (Pathak & KC, 2025). However, Pathak & Prajapati (2023) conclude that despite the presence of misinformation during the campaigning period, fears that misinformation could undermine the integrity of election results, discourage voting, or foster disinterest in the electoral system proved unfounded.

Nepal’s key interventions against misinformation are fact-checking and media and information literacy campaigns. The establishment of South Asia Check in 2015 and later initiatives such as NepalFactCheck.org (2020) and NepalCheck.org (2022) are attempts to institutionalize fact-checking. These initiatives have contributed to building databases of prominent misinformation narratives and raised awareness on the importance of fact-checking. TechPana, a Nepalese media institution, also began institutionalizing fact-checking in 2024. However, fact-checking faces limitations such as a lack of funding, which led to the closure of South Asia Check in 2023. Further issues are limited human resources, absence of technological tools tailored to local languages and a relatively low public reach (Haque et al., 2022). Operational experience from the 2026 elections highlights additional technical constraints: fact-checkers reported having access only to a limited number of AI detection tools, with no adequate tool available for certain types of content such as edited audio. Furthermore, they reported encountering numerous additional difficulties with platform-specific tracking of misinformation on TikTok, where, for instance, deleted posts could not be retraced. The absence of systematic content archiving practices and easily available tools remains also a concern. Importantly, fact-checking by nature remains reactive and slower than the spread of misinformation; therefore, its effectiveness remains low during fast-moving political events such as elections.

Media and Information Literacy (MIL) has emerged as a crucial long-term intervention to develop public resilience against misinformation (Kunwar & Prajapati, 2025). However, there is a general lack of media and information literacy among Nepali citizens (Pathak & KC, 2025). Aryal (2023) discussed media literacy as a crucial factor in building trust in media institutions and argued that despite ongoing discussions on media literacy for over a decade, public awareness and understanding of how to navigate different media platforms remain low. Expanding digital and media literacy initiatives is also necessary to equip citizens with the skills to critically engage with information (BBC Media Action, 2025). Various entities, including state, public, private and non-profit, have implemented public awareness campaigns against misinformation and some have organized training programs targeting youth and journalists, aiming to improve critical thinking and the ability to identify misleading content. Kunwar & Prajapati (2025) suggest that such programs can enhance awareness and analytical skills, but their reach remains limited and they are not yet integrated into formal education systems. Moreover, MIL programs remain scattered and have not yet reached a large population in Nepal.

At the policy level, regulatory responses have been inconsistent. While there have been attempts to bring legal frameworks to address misinformation, the process has been opaque, and proposed provisions have raised concerns regarding the state’s overreach and threats to freedom of expression. The absence of rights-based regulatory approaches means that misinformation is identified as a problem but governed haphazardly and therefore ineffectively.

An additional challenging factor is that interventions to combat misinformation are not coordinated. Surveys indicate that the general public’s responsibility for combating misinformation is shared among media, government, platforms, civil society and users (CMR-Nepal, 2022). But interventions are implemented with limited collaboration and there is no national strategy.

The role of political actors is also questionable. Political parties in Nepal have increasingly adopted social media as a primary communication tool, using dedicated digital teams to disseminate campaign messages, mobilize supporters and shape public narratives (Acharya & KC, 2025). The digitalization of political communication has also contributed to the spread of misinformation. Evidence shows that political actors and their networks have used social media to circulate misleading content, including fake screenshots of news outlets, edited videos and unverified allegations targeting opponents (Dahal, 2025).

Moreover, social media platforms have enabled direct, unmediated communication between political actors and citizens, reducing the gatekeeping role of traditional media. This change has facilitated the amplification of partisan narratives and emotional content, forms of messaging which are frequently based on misinformation. Female politicians have been disproportionately targeted by gendered misinformation campaigns calling their qualifications and private life into question. Research by the Purple Foundation (2022) shows that misinformation targeting women politicians is not incidental but actively constructed and amplified online, contributing to gendered harassment, character assassination, and broader patterns of gendered online violence in politics.

Studies on digital campaign strategies during Nepal’s local elections indicate that political parties increasingly rely on data-driven and platform-specific strategies to engage voters, including targeted messaging and coordinated content dissemination. This creates favourable conditions for organized information manipulation.

Overall, Nepal’s experience demonstrates that although multiple interventions to combat misinformation exist, they remain insufficient in scale, coordination and effectiveness. The spread of misinformation during election cycles shows challenges, including weak institutional capacity, limited resources, low media literacy and the active involvement of political actors in information manipulation. BBC Media Action (2025) emphasizes the need for a coordinated, multi-stakeholder coalition to promote information integrity and ensure collective action against misinformation. Therefore, addressing misinformation in Nepal requires strengthening existing interventions and developing a coordinated multi-stakeholder approach integrating technological, institutional and regulatory responses.

3. Findings

The findings in this section are based on the analysis of two datasets. The first dataset consists of 102 fact-checked items published by NepalFactCheck and Techpana FactCheck between February 5 and March 5, 2026, the final month of campaigning up to the election day. These items have been verified by fact-checkers and can be considered confirmed misinformation.

The second dataset consists of 1,020 entries recorded by misinformation monitors across all seven provinces of Nepal during the same time frame. The monitors are trained to identify potentially false or misleading content, but the items they flagged were not subjected to the full fact-checking process. The researchers have also reviewed all the entries. This dataset is considered potential misinformation, which is useful to understand and interpret the overall misinformation environment.

The two datasets are complementary for this study, as the fact-checked data provides depth and confirms specific cases, while the monitoring data provides a wider view of what was circulating across platforms and provinces during the pre-election period. All entries of both datasets are relevant for the general elections of 2026.

3.1 Analysis of Confirmed Misinformation

The overwhelming majority of election misinformation originated from social media, with Facebook being the dominant platform. Of the 102 misinformation pieces analysed, 87.25% were sourced from Facebook, 2.94% from TikTok, 5.88% from online media outlets, 2.94% from mainstream media outlets and 0.98% from other sources. This distribution shows that Facebook remained the primary channel through which election misinformation reached the public during the 2026 Nepal elections. A notable aspect is that even if the misinformation was originally published elsewhere, such as on newspapers, radio or television, it was subsequently circulated and spread through Facebook.

The majority of misinformation pieces was classified by fact-checkers as misleading information. This indicates that the real content had been edited, stripped of context, or circulated with a false framing. A smaller but significant amount of misinformation was classified as false or fake, which indicates completely fabricated or synthetically generated content. A few items were categorized as unverified, meaning fact-checkers could neither confirm nor conclusively disprove the claim. A small portion, 2.94%, of viral information was classified as accurate, which indicates that there was confusion among the public about what constituted accurate content. This indicates that the dominant type of election misinformation in this period involved the manipulation of real content rather than the fabrication of entirely fictional content.

The dominant actors identified as the source of such claims are content creators, especially social media authors or media producers, rather than politicians or political institutions. Only 3.92% of fact-checks attributed the original claim to a politician or a relevant expert. This points to a pattern in which misinformation was primarily produced and circulated by anonymous or pseudo-named social media accounts and content creators, rather than directly by political parties or their official channels, although the content in most cases served clear partisan purposes.

3.2 Misinformation Trends During Nepal Elections 2026

Several trends of misinformation were observed during Nepal’s 2026 elections. The misinformation items observed during the campaign period were not uniform in form or objective, as they combined different techniques such as context manipulation and edited visuals with AI-generated media and electoral fraud narratives. Most misinformation was spread to defame political leaders, but there were also other narratives that incited distrust, social polarization and uncertainty about the election. The trends identified in this study indicate that election misinformation in Nepal plays on public emotions using visually driven content that is spread primarily through social media platforms. The following sections outline the dominant patterns of the misinformation environment during the 2026 elections.

3.2.1 Most Misinformation was not Fake; it was Manipulated Context

The dominant pattern in Nepal’s 2026 election misinformation landscape was not completely fabricated content but the manipulation of authentic material. Most fact-checked claims involved real videos, speeches, photographs, or events that were stripped of context, selectively edited, or used with misleading captions. Politicians’ speeches were shortened to remove clarifying portions, old footage was circulated and unrelated visuals were used to support false narratives. Old videos were circulated that claimed to show current political events. In addition, footage of events that had nothing to do with the parliamentary elections in Nepal – from Nepal itself or even from other countries – was circulated as evidence of electoral violence in Nepal. This trend indicates that “cheapfakes” and context-based deceptions were common and far more widespread than sophisticated fabrications. This type of content gained traction because it used public figures and real visuals, indicating that election misinformation increasingly relies on cheap manipulation rather than complete fabrication, making it harder for citizens to detect falsehood.

As the production of such ‘cheapfakes’ and context-based deceptions require only limited technical know-how, little time and minimal resources, such content can be easily scaled and mass-disseminated. In conjunction with the trend whereby many social media users consume content in ever-increasing volumes, rapidly scrolling from one post to the next, such “cheapfakes” tend to be as effective as more sophisticated pieces of misinformation, yet require significantly less time and no money to generate. This combination of low technical requirements and the minimal time needed to produce “cheapfakes”, whilst maintaining the same reach among large sections of social media users, raises the worrying question of whether ever-increasing amounts of misinformation are fed to social media users in Nepal. Such a deluge of misinformation is already today straining fact-checkers and content moderators in Nepal to their breaking point. Underfunded and understaffed, their efforts to combat misinformation are increasingly proving to be a drop in the ocean in the midst of a veritable flood of misinformation.

The higher amount of cheapfakes is not only a result of their low production cost but is also due to gaps in the verification process. The absence of easily available geolocation and source-tracking tools for video content, the difficulty in verifying screenshots and deleted posts and the challenge in confirming whether content is altered create a situation in which manipulated materials can circulate largely unchecked. This means cheapfakes become dominant not only because they are easy to create but also because the tools and capacity needed to debunk them in time is not adequately available for Nepal’s fact-checkers. As misinformation is thus gradually becoming the norm rather than the exception, this could lead to a growing loss of trust in journalism as a whole. In such a scenario, it is bias and perceptions rather than critical thinking that determines whether users perceive content as genuine or fake. This could provide anti-democratic forces with fertile ground for coordinated misinformation campaigns aimed at undermining public trust in democratic institutions.

3.2.2 Election Misinformation was Leader-Centric

Election misinformation heavily targeted a small number of prominent political leaders and was not related to policy issues, party activities or election manifestos. Popular leaders such as Balen Shah, Gagan Thapa, Rabi Lamichhane and KP Sharma Oli were primary targets of election misinformation. The false claims attempted to defame the leaders by damaging their credibility as they were portrayed as corrupt, hypocritical, foreign-guided and anti-nationalist. False quotes, fake endorsements, manipulated videos and edited images were used to construct emotionally appealing narratives for the public. This trend also shows that Nepal’s digital political discourses are based on leader images, personality and emotional appeal rather than ideological and issue-based debates. The very high numbers of misinformation around popular leaders suggest that political content is gaining traction from emotional reaction and fan culture.

Such misinformation campaigns targeting political leaders reflect the general trend of a leader-centric political system in Nepal. Since the fall of the monarchy, politics in Nepal has revolved primarily around charismatic leaders rather than democratic institutions and parties. Party manifestos are not widely read, and policy issues are mostly too complex to attract widespread interest on social media. This facilitates the spread of misinformation, as algorithms favour personalization, scandals and rumours surrounding well-known political figures to a much greater extent than they would with more abstract entities such as political parties and institutions.

The concentration of misinformation around prominent male leaders appears to be driven primarily by electoral competition, particularly due to high-stakes elections where the contest for national leadership was fierce and political parties were driven by their leaders, rather than by a policy-based agenda. In this context, misinformation became a tool to damage the prospects of opponents. However, the situation differed for female politicians. Misinformation targeting female leaders mostly did not attack their political positions but focused on their private life, family relationships, behavioural characteristics and social status. In Nepal’s multicultural, multi-ethnic and multi-religious society, this gendered leader-centric misinformation carries risks that extend beyond the electoral cycle. Such misinformation, which targets women politicians, risks reinforcing a patriarchal societal perception that women in politics are reliant on male support to act as independent politicians, which ultimately undermines women’s political participation in general. Additionally, as political leaders are tied to ethnic, religious, linguistic, or caste-based identities, attacks constructed around these elements can be interpreted as attacks on the communities they represent, thereby deepening ethnic divisions and fostering a scenario of ‘us versus them’ politics.

3.2.3 AI-Generated Content is Emerging but not Dominant

During the lead-up to the elections, there were concerns about the use of AI-generated content to undermine electoral integrity and impact results (Online Khabar, 2026) and that AI-generated content is increasingly shaping the political discourse (Kantipur, 2025). This study finds that there was a growing presence of AI-generated posts; however, such content was not the dominant form of misinformation. Several fact-checks identified AI-generated images where politicians were falsely shown in hospital scenes, controversial social settings, or politically sensitive situations. AI-generated or manipulated audio clips falsely attributed statements to domestic and international political figures. Celebrity endorsement videos were altered using synthetic voice techniques. However, AI-generated misinformation remained less common than contextual manipulation and edited media. This suggests that while generative AI tools are being used in Nepal’s misinformation environment, the misinformation campaigns are primarily driven by low-cost manipulation techniques rather than technically advanced deepfakes.

The fact that context-based manipulation and edited misinformation continue to be far more common than AI-generated fakes suggests that Nepal’s misinformation ecosystem is still geared more towards altering reality rather than fabricating it. Context-based manipulations are not only cheaper but also appear to be more convincing and have a better chance of going viral than most fabricated AI content. Distorting real events by altering a headline, the date or the context still seems to be regarded by misinformation agents as more effective than completely fabricated AI-generated posts. AI-generated content also often lacks a connection to the real world. Context-manipulated misinformation, on the other hand, is linked to real events, which increases its credibility and spread.

Furthermore, manipulated audio content posed a significant challenge for fact-checkers in the run-up to the elections. Here, it is important to distinguish between two different forms of audio manipulation, which could not be adequately separated in the data underlying this study. AI-generated audio and selectively edited audio differ significantly for fact-checkers because they need to be identified differently. Completely synthetic audio recordings, in most cases, can be quickly identified, but audio recordings that have been manipulated to distort the meaning are difficult to verify and debunk; this requires more time and resources and, in some cases, cannot be verified at all. This distinction is important because it directly impacts how fact-checkers operate, as such audio recordings require more time and resources. This goes beyond the general unavailability of AI detection tools.

Another explanation for the relatively limited use of AI-generated content might be that AI-generated or manipulated audio clips in Nepali tend to be less sophisticated than those in languages such as English, resulting in clumsy phrasing and unnatural speech patterns that the target audience can easily detect. This is due to the still comparatively low volume and quality of AI training data for languages such as Nepali, which makes AI systems significantly less capable and therefore more difficult to use for AI-generated misinformation. At the same time, AI systems for the most widely spoken languages in Nepal, such as Nepali, Maithili and Bhojpuri, are improving rapidly. The fact that AI-generated content is already being used for misinformation in the 2026 election cycle highlights the rise of such content, the prevalence of which is likely to increase in Nepal’s misinformation landscape as technology progresses and applications become more affordable and user-friendly.

The gap between the AI-generated content identified in the monitoring dataset (23%) and in the confirmed fact-check dataset is not solely due to the lack of detection tools; access to these tools also poses a genuine problem. It also shows the challenge that content deleted from the social media platforms becomes very difficult to verify – or in many cases impossible to verify at all – even if the monitors have recorded it. Posts that are deleted before they can be fact-checked are an issue that cannot be resolved simply by ensuring the availability of detection tools. Many of the archiving tools do not store social media posts and, even when they do, they are not reliably accessible to the public.

3.2.4 Violence and Fraud Narratives were used to Undermine Electoral Integrity

Another major trend in Nepal’s 2026 election misinformation landscape was the circulation of narratives related to violence, electoral fraud and uncertainty. False claims of attacks on politicians, gunfire, booth capture, vote buying and election-related clashes appeared repeatedly, especially in the final week leading to the election day. Many of these claims used unrelated videos, old footage or misleading captions to show electoral chaos and insecurity. Police firing, armed confrontation and voter intimidation were dominant themes of these misinformation pieces. Their timing and nature showed that they were created not to favour or damage candidates but to undermine public confidence in the fairness and security of the election. This type of misinformation aimed to negatively impact public trust in the election process during a fragile political transition.

Such misinformation narratives may have served a variety of different objectives in the context of Nepal’s elections. As the violent events surrounding the Gen Z protests were still fresh in the minds of Nepalese people across the country, misinformation about violence at polling stations may have been used to deter voter turnout in closely contested constituencies due to the perceived physical danger. At the same time, in such a volatile post-transition electoral environment, fabricated claims of foreign interference could have served to delegitimise the entire electoral process and justify future protests against those who have emerged victorious by portraying them as illegitimate winners of a rigged election.

This type of misinformation narrative rose to prominence elsewhere during the 2020 United States presidential election, when claims were made that the election had been rigged. Fabricated accounts of such electoral fraud and manipulation were spread via social media and later reinforced by legal challenges and media statements by the then-defeated candidate and present-day US President Donald Trump, despite numerous court rulings finding no evidence of fraud. To this day, despite numerous fact-checks, this narrative remains widespread among large sections of the American population.

Similarly, in the case of the 2026 elections in Nepal, the high prevalence of such misinformation narratives could have been used by some political actors as both a pre- and post-election strategy. Firstly, to delegitimise political opponents and secondly, to undermine public confidence in democratic processes within an already highly volatile political environment. The success of the spread of such misinformation narratives can be attributed to the already high level of tension surrounding a fragile political transition, in which emotions such as fear and mistrust towards political parties and democratic institutions were particularly pronounced among many Nepalis. Consequently, emotional claims of electoral fraud and violence found a receptive audience, with fear and anxiety often spreading faster than official statements could counter these fabricated claims.

Field reporting during the election period observed that many voters were receiving important information primarily through social media rather than television or print and they were treating content they saw on their mobile phones as factual when making electoral decisions. Although the direct impact on voter behaviour cannot be quantified in this study, such observations suggest that the information environment in which these narratives circulated was precisely the place where citizens had been actively forming their political opinions. One further point that needs to be considered behind the rapid spread of misinformation narratives is the monetisation of social media content, which creates incentives to disseminate sensational and unverified claims. The decision by Meta to monetize Facebook ahead of the elections surged the number of such content, which was spread not only by influencers but also by journalists and others. This shows that the circulation of false narratives was not only politically motivated but also linked to financial gains.

3.2.5 Visual Misinformation Dominated Social Media

Visual content – either photos or videos – was central to the spread of election misinformation across social media platforms during the election of 2026. Manipulated photos, edited videos, fake screenshots and misleading visual clips formed most viral misinformation pieces identified during the study period. Facebook was the primary dissemination platform, where visual content circulated quickly through engagement, as well as specific pages and closed groups. Visual misinformation was effective because it can be consumed quickly, inciting emotional responses and is often taken as evidence. The prevalent use of edited images, recycled videos and AI-generated visuals also shows a trend of political misinformation being image-based. This is the case, because images are prioritized by social media algorithms, suitable across multiple digital platforms and easy to consume through mobile phones. This trend indicates that political narratives are shaped through short-form and emotionally engaging visual media pieces rather than detailed textual arguments.

This trend, whereby visual content serves as the preferred medium for misinformation, could place an increasing burden on fact-checking systems, as the dissemination of such content is highly scalable. Whilst manipulated visual content can be quickly shared across various platforms and captioned in multiple languages, the fact-checking process is increasingly slowed down. This is particularly true for videos, which require frame-by-frame analysis for verification. In the Nepalese context, with its multilingual population, image-based misinformation also bridges linguistic boundaries and thus potentially reaches a much wider demographic, resulting in an increasingly growing advantage for misinformation over verification mechanisms. At the same time, this trend is reinforced by the algorithmic architecture of social media platforms, with social networks such as Facebook prioritising highly emotional visual content over lengthy text-based content.

3.3 Analysis of Potential Misinformation

Provincial misinformation monitors across all seven provinces of Nepal recorded 1,020 instances of potential election-related misinformation between February 5 and March 5, 2026. This monitoring data provides a broader picture of the misinformation environment as it contains content flagged as potentially false or misleading. All the instances were not verified to the standard required for a published fact-check, thereby limiting the researchers’ ability to confirm them as misinformation. The daily average of flagged content was 24 items in the first half of the monitoring period, which increased to 44 in the final 15 days before the election. This volume and increment is consistent with the pattern observed in the fact-checked data and shows that the number of misinformation pieces increased significantly as voting day neared.

3.3.1 Platforms of Misinformation Dissemination

Social media accounted for 87 percent of the 1,020 pieces of potential election-related misinformation, with online media contributing 9 percent and social media-based pages or groups – accounts that present themselves as news outlets but operate without editorial standards – accounting for over 3 percent. Television and speeches together made up less than 1 percent.

Within social media platforms, Facebook dominated, accounting for 88 percent of social media instances, followed by TikTok at 9 percent and Twitter/X at under 2 percent. YouTube accounted for less than half a percent of monitored potential misinformation pieces.

An important observation is that even when misinformation originated in mainstream media, online outlets, or public speeches, it was primarily consumed and amplified through social media platforms. This means that the platform distribution figures likely overstate Facebook’s role in the misinformation ecosystem, since content produced elsewhere was also redistributed there to reach a larger audience.

Facebook remains the social media platform with the largest user base and the widest cross-generational audience in Nepal. As it is Nepal’s largest public gathering place for everything – from politics and news to the job market, as well as a cross-border communication platform through which many Nepali migrant workers stay in touch with their families – it is not surprising that most misinformation would end up on Facebook in one way or another. In this role as the most popular social platform with the largest user base, misinformation tends to go viral more easily than on other emerging social media platforms, such as Twitter/X and TikTok. False information shared by a trusted family member or a respected community leader is often forwarded without verification, because trust in the content rests not on verification mechanisms but on social and emotional ties. As likes and shares increase, such misinformation gains greater emotional credibility, leading to further dissemination and increased virality. Facebook’s limited moderation capacity for content in Nepali or other regional or local languages could have exacerbated the issue, as content that would typically be flagged quickly in Western markets often circulates for much longer in Nepali or other regional languages, sometimes without being flagged at all.

3.3.2 Types of Contents

The content-type breakdown shows that texts were the single most common form, accounting for 27 percent of flagged items, followed by edited videos at 24 percent and images at 23 percent. Text-on-image, a format in which a claim or quote is overlaid on a photo, accounted for 13 percent. Animation, raw unedited videos and memes made up the remainder.

The dominance of edited videos, constituting nearly a quarter of all flagged content, is consistent with the pattern observed in the fact-check data, where context manipulation through altered or misrepresented visuals proved to be the most common method. The combination of text, image and text-on-image formats, which together account for over 60 percent of flagged content, also shows the low production cost of election misinformation. These formats require no specialist tools and can be produced and shared rapidly. Misinformation is thus democratised, as anyone possessing a smartphone can create and disseminate such content. In the Nepalese context, misinformation seems not to depend on technical refinement, but rather on velocity, volume and emotive appeal to reach its target audiences. This confronts fact-checkers and content moderators with an ever-growing flood of misinformation, the scale of which far exceeds their capacity in terms of time, resources and coordination capabilities.

3.3.3 Actors of Misinformation

Influencers as well as individuals with established social media followings who produce and share political content were the primary misinformation actors, contributing to 40 percent of all flagged items. Social media pages or groups accounted for 20 percent and political figures, including party leaders and cadres, for 16 percent. Journalists or media personnel accounted for nearly 12 percent and anonymous accounts for under 10 percent.

These figures indicate that the majority of flagged misinformation was produced or amplified not by anonymous actors or formal political party structures, but by individuals and accounts with identifiable public presences such as influencers, media figures and social media pages. Their large audiences gave the misinformation content reach and credibility. The relatively low number of anonymous accounts involved in spreading misinformation challenges a global common assumption that election misinformation is primarily driven by unknown people. However, this finding should be interpreted with caution. During the election period, a significant number of social media channels were found to be presenting themselves as news media outlets, even though they were not. The use of impersonating accounts created using another person’s name, photograph, or identity, as well as fake accounts that use non-existent names, personal information and photos, were also observed. This all added to public confusion about the distinction between valid sources and individual or fictitious pages, as well as anonymous accounts giving the impression of being personal accounts. Such accounts were categorised as influencers or social media pages rather than anonymous accounts, which means that the role of fake identity construction in Nepal’s misinformation ecosystem may have been understated in this study.

The fact that most of these influencers and popular digital content creators are not officially affiliated with any political party or ideology increases the credibility of the information they share. Combined with their extensive networks and emotional sway, the misinformation they disseminate appears to reach a far greater number of people with a broader range of political beliefs than would be the case, for example, with partisan political actors. This growing role of influencers as key players in the spread of political misinformation reflects a broader transformation in Nepal’s current political communication landscape. It points to a massive shift in influence over political discourse away from traditional mainstream media outlets and trained journalists towards social media personalities on algorithm-driven platforms. This trend is dangerous, as the content produced by most of these influencers is shaped by audience engagement numbers rather than by universal journalistic standards and ethics. This, in turn, is contributing to the increasing commercialisation of political discourse in the country. As social media platforms tend to reward attention, interaction and emotion rather than neutral reporting, sharing highly controversial political content and misinformation becomes more profitable, as it generates more reactions, views and, consequently, revenue.

At the same time, the relatively high rate of 12 percent of journalists and media professionals disseminating misinformation points to systemic issues within the Nepalese media landscape. In addition to political pressure and bias, this high rate might also indicate structural challenges of journalism in the country. In the face of shrinking budgets and limited resources, fact-checking seems to have become a secondary concern compared to the race for breaking news. These findings may also point to gaps in professional standards, as many Nepalese journalists appear to lack training in editorial ethics or the practice of adhering to them. In such an environment of information disorder, where even professional journalists and mainstream media platforms are deeply involved in the dissemination of misinformation, the credibility of the entire information landscape could be undermined. This could further erode the already low level of trust among the Nepalese public in journalism, institutional bodies and the mainstream media in general.

3.3.4 Use of AI in Misinformation

Twenty-three percent of the 1,020 flagged items during the 2026 election in Nepal were identified as AI-generated content, which means that nearly one in four misinformation pieces used synthetic or AI-produced material.

This is higher than what the fact-checked misinformation showed. This indicates that AI-generated content has become common; however, the fact-checkers lack tools and expertise in verifying these contents. This disparity shows two problems. First, fact-checkers currently rely on a limited set of AI detection tools that are insufficient to properly identify a high number of AI-manipulated content, particularly for selectively edited audio recordings. Second, content that is flagged by monitors but subsequently deleted from social media platforms becomes very difficult to verify retrospectively, meaning that some AI-generated items in the monitoring dataset may remain unconfirmed, not because they are not misinformation, but because they have been removed. Therefore, the data indicates that AI-generated content has become more common in Nepal’s misinformation ecosystem, but the capacity to verify it has not kept pace.

That only 0.4 percent of flagged content was identified as sponsored or paid is also a notable aspect. It suggests that the amplification of election misinformation during this period was driven primarily by organic sharing through influencer networks, party supporter communities and social media pages rather than through paid promotion. The volume and speed of misinformation distribution appear to have been achieved without significant financial investment in boosting posts but rather relied on the existing networks of users who share content that confirms their political biases. The pattern suggests that misinformation in Nepal’s social media landscape is structurally embedded in political bias, social trust networks and platform designs, rather than being driven by top-down coordination. The spread of misinformation could therefore be described as both social and algorithmic, which makes its dissemination potentially far more dangerous. Its organic sharing implies that most misinformation pieces do align with the perceptions, narratives and prejudices shared by large sections of the population and thus are socially validated. At the same time, this organic spread of misinformation suggests that the risk of inciting violence in the real world is much higher, as such misinformation pieces amplify hostilities, hatred and prejudices that are already deeply ingrained in society. This organic spread of misinformation is also motivated by the monetisation incentives of social media platforms, especially those of Meta, YouTube and TikTok, which reward viral and emotionally engaging content regardless of its accuracy, creating an environment in which the spread of unverified content is economically beneficial and socially normalised.

3.4 Comparison with Nepal’s 2022 Elections

This study also used similar data from the 2022 Nepal elections for comparison. The 2022 dataset includes fact-checks published during a period of a month leading up to the polling days, i.e. May 13, 2022, and November 20, 2022. In 2022, NepalFactCheck.org published 24 election-related fact-checks across two election events, the local level election in May and the federal and provincial elections in November. The 2026 dataset contains 102 items from a single election cycle, produced by two fact-checking platforms. Numerically, the fact-checking capacity doubled by the addition of Techpana FactCheck; the rise in volume indicates a significant increase in the spread of verifiable misinformation during the 2026 cycle.

In 2022, the volume of misinformation was concentrated: most false claims appeared in the days immediately before polling. In 2026, misinformation was present throughout the final month of campaigning, which increased in frequency as election day approached.

The dominant type of political misinformation also changed between the two elections. In 2022, the most common verdict was outright false – claims that were factually incorrect, such as a fabricated bomb blast report or a fake BBC survey. In 2026, the dominant verdict is misleading, pointing to a shift towards contextual manipulation, with real videos and images reused with false labels, speeches clipped to distort meaning and photos edited to change what they showed. This shift matters for how misinformation is detected and countered. A fabricated claim can be debunked by establishing that an event did not occur. A misleading claim requires establishing what actually happened, recovering the original context and demonstrating how the presentation altered the meaning – a more resource-intensive process that takes longer, thereby losing necessary time to reach the public, for whom the misinformation is still relevant.

The use of AI-generated content shows a big change. In 2022, there were no documented cases of AI-generated images, synthetic audio or deepfake videos. In 2026, there is a significant amount of misinformation involving content identified using AI detection tools, and several pieces of misinformation were produced using multiple manipulations, such as authentic footage with fabricated audio or photographs with faces or objects digitally replaced. As AI tools for Nepali, Maithili and Bhojpuri languages are advancing rapidly and the relevant applications are becoming increasingly user-friendly and cost-free, the trend towards the growing use of AI to spread misinformation is likely to intensify further. The use of such tools means that misinformation can spread even more rapidly and becomes increasingly difficult to verify. Traditional verification methods quickly become outdated, as fact-checkers must constantly unlearn old practices and acquire new ones to keep pace with technological progress. As a result, fact-checkers already have to rely increasingly on AI tools to identify AI-generated content.

One continuity between 2022 and 2026 is the concentration of misinformation around high-profile candidates. In 2022, Rabi Lamichhane was the subject of a false claim about vote-counting results in Chitwan. In 2026, Lamichhane appeared in multiple fact-checks in claims involving AI-generated images, misrepresented diplomatic meetings and fabricated political agreements. Similarly, the targeting of Balen Shah and Gagan Thapa, both of whom had become nationally prominent between 2022 and 2026, follows the same path seen in 2022. The narratives used hospitalization, crowd rejection, community insults and secret deals to attack the reputation of prominent leaders. This focus on prominent party figures is not a phenomenon specific to Nepal’s online misinformation landscape but rather reflects the country’s political culture. Most Nepalis tend to look to charismatic leaders rather than their parties or democratic institutions. This has been picked up by actors in the realm of misinformation, who appear to have mimicked these patterns to increase the reach of their misinformation pieces.

A second continuity is the role of Facebook as the primary platform for the circulation of election misinformation. In 2022, the fact-check record did not systematically document platform-level data, but the cases predominantly referenced content circulating on Facebook. In 2026, 89 of the 102 items in this dataset originated from Facebook. This suggests that despite changes in the social media landscape and growing attention to TikTok as a misinformation vector, Facebook remained the central platform through which false election content reached the largest audience in Nepal. If Facebook retains this dominant position, it can be expected to remain the most widely used platform for the spread of misinformation. Compared to TikTok, which has the second-largest user base in Nepal, Facebook’s usage patterns also seem to facilitate the spread of political misinformation to a far greater extent, as the platform is mainly used for socialising, networking and as a news platform, so political discussions tend to take place in this space. TikTok exhibits a very different usage pattern in Nepal, focusing mainly on short entertainment videos such as dances, comedy and music and to a much lesser extent on political discussions.

However, this finding should not underestimate TikTok’s role in the misinformation ecosystem. TikTok’s audio-reuse feature makes it riskier for misuse of emotional content, with videos depicting deaths, accidents or dramatic events ready to be seamlessly re-captioned and recirculated to mislead. TikTok also emerged as a dominant vector for AI-generated content targeting women politicians, including videos depicting female leaders in romantic, intimate or sexualised scenarios, which circulated widely during the election period. These differing usage patterns, should they persist, also suggest that Facebook could remain the central platform for the spread of misinformation in Nepal, even as TikTok functions as a distinct and growing vector for emotional and gendered misinformation content.

4. Conclusions and Recommendations

Nepal’s 2026 general elections were held in a fragile information environment. The rise of misinformation has threatened to disrupt the elections by making the campaign period messy, inflaming identity sentiments, inciting violence and altering the results. The dissolution of parliament, the formation of an interim government, the political changes due to the Gen Z movement and declining public trust in both media and state institutions had created a situation which proved fertile ground for misinformation. This study shows that there were attempts to exploit the situation by spreading misinformation for political gains, either to promote a certain party, candidate, or narrative, or by attacking the opposition.

4.1 Conclusions

The major findings of this study show a complex misinformation landscape in Nepal.

First, the dominant form of election misinformation in 2026 was manipulation, not fabrication. Unlike the 2022 elections, during which outright false claims were more common, the 2026 elections saw more out-of-context and misleading content. Real speeches were edited, changing their meaning; old videos were circulated as new; and photos were digitally manipulated to add false elements. This “cheapfake” manipulation has implications for fact-checking, which requires not only debunking the misinformation but also finding the original context or incident.

Second, AI-generated content has become more common in Nepal’s misinformation landscape, though it has not yet become dominant. Synthetic audio, fabricated imagery and voice-cloned narration over real footage were observed. The difference between monitored prevalence and fact-checks indicates a capacity gap in countering AI-generated misinformation. This gap, however, has two dimensions. Fact-checking organizations in Nepal currently lack adequate tools and skills to verify AI-generated content, particularly for edited audio, which is much more difficult to verify than completely synthetic audio clips. In addition, content that is deleted from platforms is often unverifiable, creating a problem that goes beyond tool availability. Addressing Nepal’s AI misinformation challenge therefore requires both investment in verification tools tailored to local language contexts and the development of archiving and documentation practices that preserve flagged content before it is removed.

Third, misinformation was concentrated around major political figures such as Balen Shah, Gagan Thapa, Rabi Lamichhane, KP Sharma Oli and Pushpa Kamal Dahal Prachanda, who were repeatedly targeted through recurring narratives such as hospitalization, attacks, corruption, secret deals, communal insults and loss of public legitimacy. Such narratives were not new, but they were still effective, as they played on pre-existing political anxieties and biases to attract emotional reactions such as anger, disgust and fear.

Fourth, misinformation was used to undermine trust in the electoral process. Narratives of booth capture, vote buying, gunfire and foreign interference increased significantly during the final days of the campaign and on polling day. Such misinformation pieces attempted to delegitimize the elections and created the perception of electoral instability. While the direct effect on voter turnout cannot be quantified from this study, field observations indicated that many voters were primarily consuming and acting on information received through social media. This indicates that the conditions for behavioural impact existed, even if the scale of that impact was not measured.

Fifth, Facebook remained the dominant distribution channel for election misinformation in Nepal. Despite increased use of TikTok and YouTube, Facebook remains the central platform for the circulation of misinformation. Misinformation spread largely through organic sharing among politically aligned groups rather than through paid advertising. This highlights the role of platform design, recommendation systems and confirmation-bias dynamics in the circulation and amplification of misinformation in Nepal.

Sixth, gendered misinformation is a distinct and structurally embedded pattern. Women in politics were targeted through fabricated sexual content, invented statements and coordinated attacks on their personal and family lives.

The findings of this study confirm that misinformation in Nepal’s 2026 elections was not random or spontaneous. It was systematically produced and distributed by identifiable actors, primarily content creators with established audiences and political affiliation, using low-cost techniques that manipulated social platform design, leading to political polarization and declining public trust in the media.

4.2 Recommendations

A. For State Institutions and the Election Commission of Nepal (ECN)

1. Establish a pre-election misinformation rapid-response protocol. The ECN should establish a dedicated misinformation monitoring unit that begins round-the-clock monitoring at least 30 days before any election. The ECN should develop a clear procedure and monitoring template well ahead and establish communication channels with social media platforms to report problematic content. Given that violence and electoral fraud narratives appeared to be coordinated around the voting day itself, the unit should prioritise real-time monitoring during polling day. The ECN should also collaborate with fact-checking institutions to share monitoring data and establish a debunking mechanism to enable false claims about the integrity of elections to be swiftly debunked.
2. Develop rights-based regulatory frameworks for platform accountability. Social media platforms have limited moderation capacity for Nepali and regional languages, which allows misinformation to circulate far longer than equivalent content in Western markets. Regulatory engagement with platforms should require minimum-standard content moderation in Nepali and regional languages, as well as transparent takedown reporting. Any such framework must be developed in consultation with civil society and guided by constitutional guarantees of freedom of expression.
3. Define misinformation as a real-world conflict indicator. The study documents how narratives targeting religious minorities, ethnic communities and caste identities carry potential for offline harm. The ECN and relevant state institutions should include digital information flows as a formal conflict-risk variable in election monitoring.

B. For Civil Society Organisations (CSOs)

1. Pair fact-checking with proactive narrative monitoring. This study showed that the dominant form of misinformation in 2026 was not outright fabrication but contextual manipulation that exploits real events, real faces and real footage. This shift makes debunking more resource-intensive and time-consuming than correcting a fabricated claim. CSOs should invest in narrative tracking that identifies recurring misinformation types, such as foreign interference, booth capture, violence and religious identity issues, before they spread, so that their effectiveness is minimized.
2. Invest in AI detection tools adapted to local language contexts. The gap between the 23% AI-generated content in the monitoring dataset and the much lower rate confirmed through fact-checks reflects a capacity gap. Fact-checking organisations currently lack adequate tools to fact-check and archive social media posts. CSOs should seek funds and invest in AI detection tools adapted for local languages to build technical capacities.
3. Develop a multi-stakeholder coalition against misinformation. Misinformation is a serious threat that requires a whole-of-society approach. CSOs should support the development of a multi-stakeholder coalition, including fact-checkers, journalists, digital rights advocates, academic researchers and community monitors, with shared protocols, joint monitoring infrastructure and coordinated public communications.
4. Expand and institutionalise gendered misinformation monitoring. The study confirms that women in politics are targeted through qualitatively different misinformation, such as fabricated sexual content, attacks on family and private life, and character assassination that extends beyond electoral cycles and has the potential to cause broader social harm. CSOs working on misinformation should establish monitoring for gendered content, build documentation that preserves evidence of such attacks and link these findings to broader advocacy on women’s political participation and online safety.
5. Include misinformation in conflict early warning. CSOs working on conflict prevention and social cohesion are well placed to monitor identity-based misinformation as part of their existing early-warning work. The study found that religious and ethnic misinformation narratives have a documented history of triggering real-world violence in Nepal. CSOs should ensure their conflict-monitoring tools also cover digital content targeting minority communities and establish collaboration with fact-checkers and state authorities when high-risk content is identified.
6. Facilitate community dialogue on misinformation. CSOs with experience in dialogue facilitation should develop community-level dialogue on the relationship between online misinformation and offline harm. These dialogues should be part of their broader peacebuilding and conflict-transformation mandate.
7. Include media and information literacy within peace education and civic programming. CSOs supporting peace education, youth civic engagement and community resilience should recognise MIL not as a separate intervention but as a vital component of peace education.

C. For Media Organisations and Journalists

1. Establish fact-checking as a standard operating procedure. The finding that nearly 12% of flagged misinformation was produced or amplified by journalists and media personnel points to a systemic failure to uphold journalistic standards. Economic pressure, shrinking newsrooms and the race for breaking news have eroded fact-checking as a standard practice. Editors and media organisations must ensure internal verification standards and recognise that spreading misinformation, even unintentionally, erodes the public trust on which journalism ultimately depends.
2. Develop platform-specific content strategies. The dominance of visual misinformation shows how platform algorithms prioritise emotionally engaging visual content. Media organisations should train journalists in reverse image search, video verification and metadata analysis, and build these skills into editorial workflows.
3. Ensure conflict-sensitive reporting, with particular focus on identity-based misinformation. The study documents how misinformation targeting religious, ethnic and caste identities has repeatedly triggered communal violence in Nepal. Media organisations should adopt conflict-sensitive reporting guidelines that address how to cover misinformation without amplifying divisive narratives and existing social stereotypes.
4. Report on misinformation as a beat. Citizens’ ability to shield themselves from misinformation depends on their awareness of how misinformation is created and disseminated. Continued reporting that explains the various interlinked aspects of misinformation helps foster media literacy among the general public.

Abbreviations

• AI=Artificial Intelligence
• BBC=British Broadcasting Corporation (as in BBC Media Action)
• CMR-Nepal=Center for Media Research – Nepal
• CSO(s)=Civil Society Organisation(s)
• ECN=Election Commission of Nepal
• Gen Z=Generation Zoomers
• IRI=International Republican Institute
• MIL=Media and Information Literacy
• NCP=Nepal Communist Party
• NDI=National Democratic Institute
• PCN=Press Council Nepal
• RPP=Rastriya Prajatantra Party
• RSF=Reporters Without Borders
• RSP=Rastriya Swatantra Party
• UML=United Marxist–Leninist

References

• Acharya, U. (2025a). Let’s dismiss the term ‘fake news’ to combat misinformation. Nepali Media: Issue of Ethics and Sustainability, 5, 1–16. nepjol.info/index.php/nmies/article/view/85892
• Acharya, U. (Ed.). (2025b). Misinformation and democracy. In Nepal’s misinformation landscape. Center for Media Research–Nepal. DOI: 10.62657/cmr25a0a
• Acharya, U. (2025c). Misinformation during the Gen Z uprising in Nepal: Trends and lessons learned for interventions. Center for Media Research–Nepal. DOI: 10.62657/cmr25opa
• Acharya, U., & Banu, S. (2025). Understanding misinformation in Nepal: Expert views on definitions, determinants, solutions, and way forward. Purak Asia. DOI: 10.62657/purak2509
• Acharya, U., & KC, B. (2025). Social media ra patrakarita [Social media and journalism]. In J. B. Shreeram Singh Basnet & D. Khanal (Eds.), Nepali mediako itihas: Chhapadekhi digitalsamma [History of Nepali media: From print to digital]. Press Council Nepal.
• Anupam, B. (2025). Nepal’s misinformation domain: Comparing mainstream journalists and content creators. Nepali Media: Issue of Ethics and Sustainability, 5, 159–172. nepjol.info/index.php/nmies/article/view/85905
• BBC Media Action. (2025). Understanding mis- and disinformation in Nepal. bbc.co.uk/mediaaction/documents/understanding-mis-and-disinformation-in-nepal.pdf
• Bennett, W. L., & Segerberg, A. (2013). The logic of connective action: Digital media and the personalization of contentious politics. Cambridge University Press.
• Bhandari, B. (2024). Weaponizing information: The rise of social media manipulation in Nepal. Journal of Durgalaxmi, 3. nepjol.info/index.php/jdl/article/view/73833
• Bhandari, P., & Desruisseaux, T. M. (2025). Digital literacy and its impact on employment in Nepal. International Journal of Asian Social Science, 15, 315. archive.aessweb.com/index.php/5007/article/view/5635
• Bhattarai, D. (2023). Combatting information disorder: A South Asian perspective. In J. M. Karen Fowler-Watt (Ed.), The Palgrave handbook of media misinformation. Springer International Publishing.
• boyd, d. (2014). It’s complicated: The social lives of networked teens. Yale University Press.
• Brashier, N. M., Pennycook, G., Berinsky, A. J., & Rand, D. G. (2021). Timing matters when correcting fake news. Proceedings of the National Academy of Sciences, 118(5), e2020043118. doi.org/10.1073/pnas.2020043118
• CMR-Nepal. (2022, May). Nepal social media users survey 2021. research.butmedia.org/wp-content/uploads/2022/06/SocialMediaSurvey_Nepal_2021_CMR.pdf
• Connors, G., & Baumhofer, E. (2025). Peacebuilding and disinformation: Taking stock and planning ahead. Berghof Foundation & Platform Peaceful Conflict Transformation. berghof-foundation.org/library/peacebuilding-and-disinformation
• Dahal, G. (2026). Gen Z protest and its prospects in strengthening democracy in Nepal. Journal of Political Science, 26, 203–216. nepjol.info/index.php/JPS/article/view/90796
• Dahal, R. (2025a). Digital paridrisya: Sanchar ra patrakaritako aayamik pariwartan [Digital landscape: Dimensional transformation of communication and journalism]. Center for Media Research–Nepal. DOI: 10.62657/cmr25b
• Dahal, R. (2025b). Framing misinformation: A review of literature and media discourse. In U. Acharya (Ed.), Nepal’s misinformation landscape. Center for Media Research–Nepal. DOI: 10.62657/cmr25ac
• Deutsche Welle. (2024, February). Press in crisis in Nepal as media revenue dips. dw.com/en/journalism-in-crisis-in-nepal-as-media-revenue-slumps/a-68397162
• Dhungel, B. P., & Gyanwali, G. P. (2025). Impact of social media on teenage girls of Nepal. National College of Computer Studies Research Journal, 4, 19. nepjol.info/index.php/nccsrj/article/view/84354
• Ecker, U. K. H., O’Reilly, Z., Reid, J. S., & Chang, E. P. (2020). The effectiveness of short-format refutational fact-checks. British Journal of Psychology, 111(1), 36–54. doi.org/10.1111/bjop.12383
• Global Issues. (2025). Nepal’s Gen Z protest: How fake news tried to rewrite a revolution. globalissues.org/news/2025/09/18/41053
• Haque, M., Yousuf, M., Arman, Z. R., Rony, M. U., Alam, A. S., Hasan, K. M., Islam, K. M. S., & Hassan, N. (2022). Fact-checking initiatives in Bangladesh, India, and Nepal: A study of user engagement and challenges. arXiv. arxiv.org/abs/1811.01806
• Kandel, N. (2020). Information disorder syndrome and its management. JNMA: Journal of the Nepal Medical Association, 58, 280–285. pmc.ncbi.nlm.nih.gov/articles/PMC7580464/
• Kantipur. (2025, December). Rajnitima AI ko badho prayog [Increasing use of AI in politics]. ekantipur.com/news/2025/12/21/…
• Kharel, A. B. (2024). Cyber-politics: Social media’s influence on political mobilization. Journal of Political Science, 24, 103–115. nepjol.info/index.php/JPS/article/view/62857
• Koirala, J. (2025, September 11). Gen Z protests and role of social media in Nepal (September 8–10, 2025) [Working paper]. Cambridge Open Engage. doi.org/10.33774/coe-2025-x0j4z
• Kunwar, C., & Prajapati, U. (2025). Media and information literacy: Examining the effectiveness of educating youth. In U. Acharya (Ed.), Nepal’s misinformation landscape. Center for Media Research–Nepal. DOI: 10.62657/cmr25ae
• National Democratic Institute, International Republican Institute, & Stanford Internet Observatory. (2024). Suchana chalkhelko mukabila: Nirbachan ra tyasko seroferoka lagi srotpustika [Combating information manipulation: A playbook].
• Nepali Times. (2009, July). More insecure. archive.nepalitimes.com/news.php?id=16103
• Onlinekhabar. (2025, January). AI and fake accounts during Nepal’s Gen Z uprising. english.onlinekhabar.com/ai-fake-account-gen-z-uprising.html
• Onlinekhabar. (2026a, January). AI ra sanjalle prababit banauna sakchha chunabi natija [AI and social networks can impact election results]. onlinekhabar.com/2026/01/1855107/…
• Onlinekhabar. (2026b, January). Protests erupt in Birgunj after TikTok video triggers religious tension. english.onlinekhabar.com/protests-erupt-in-birgunj.html
• Panos South Asia. (2022). Analysis of gendered violence in media against women in politics in Nepal.
• Pathak, T., & KC, B. (2025). Misinformation: Shared concern, limited interventions. In U. Acharya (Ed.), Nepal’s misinformation landscape. Center for Media Research–Nepal. DOI: 10.62657/cmr25a0
• Pathak, T., & Prajapati, U. (2023). Online platforms, polling, and misinformation: Electioneering strategies in Nepal. influenceindustry.org/en/explorer/case-studies/nepal-electioneering-strategies/
• Prajapati, U., & Pandey, L. (2025). Public perceptions of misinformation: Sources, impacts and societal implications. In U. Acharya (Ed.), Nepal’s misinformation landscape. Center for Media Research–Nepal. DOI: 10.62657/cmr25ab
• Purple Foundation. (2022). Gendered violence against women and gender and sexual minorities in politics in Nepal. purplefoundation.org.np/uploads/publications/download-pdf-1678955318.pdf
• Ray, A. (2025). How misinformation fuelled panic during Gen Z uprising. The Kathmandu Post. kathmandupost.com/national/2025/09/15/…
• Roy, A. (2025). Report on Nepal’s Gen Z protest comes under scrutiny. The Kathmandu Post. kathmandupost.com/national/2025/10/12/…
• Rupakheti, B. K. (2025). Social media’s role in shaping political news consumption in Nepal. NPRC Journal of Multidisciplinary Research, 2, 86. nepjol.info/index.php/nprcjmr/article/view/83835
• Sapkota, M. (2026). Gen Z movement 2025 in Nepal: Agendas, implications and unsettled issues. Journal of Political Science, 26, 144–173. nepjol.info/index.php/JPS/article/view/90793
• Sharma, B. P. (2025). Increasing distrust in mainstream media and the state of the online and social media in Nepal. International Journal of Media, Journalism and Mass Communications, 11, 49–60. arcjournals.org/pdfs/ijmjmc/v11-i1/5.pdf
• The Asia Foundation. (2024). Vibrant information barometer: National report—Nepal. asiafoundation.org/wp-content/uploads/2024/09/…
• The Kathmandu Post. (2023, October). Curfew imposed in Nepalgunj over fears of communal violence. kathmandupost.com/national/2023/10/03/…
• Timilsina, K. P. (2026). Nepal’s political trajectory and challenges to democratic stability. Journal of Political Science, 26, 189. nepjol.info/index.php/JPS/article/view/90795
• Vasist, P. N., Chatterjee, D., & Krishnan, S. (2024). The polarizing impact of political disinformation and hate speech: A cross-country configural narrative. Information Systems Frontiers, 26, 663–688. pmc.ncbi.nlm.nih.gov/articles/PMC10106894/

Credits & Acknowledgements

We would like to thank all those who supported this study.