A Nepal survey found awareness and digital skills tracked with spotting fake video, but it measured what 400 people believed, not what stopped a fake.

survey · Nepal Journal of Multidisciplinary Research · la publicación, 1 jul 2026 · gratis

En Nepal, la gente cree que las campañas de información y la educación ayudan a frenar los videos falsos.

Versión breve · la versión detallada sigue, unos 6 min

Pregunte al chat de weeklyAI

Pregúnteme por este estudio: a quiénes se estudió, qué encontró y qué no dice.

Las conversaciones se guardan mientras exista weeklyAI, para mejorar la publicación. Se responde en el idioma en que usted escribe.

El estudio, de un vistazo
Quiénes: personas que usan redes sociales en Nepal
Cuántos: 400
Dónde: Nepal
Cuándo: entre el 3 de septiembre de 2025 y el 21 de enero de 2026
Tipo de estudio: encuesta
Quién lo hizo: University of Wolverhampton y Herald College Kathmandu
El límite que importa: Solo mide lo que la gente cree, no si de verdad se detectan o se frenan los videos falsos.

La pregunta que le queda no es si el video es falso, sino a quién puede pedirle el informe que lo confirma.
Lectura de weeklyAI

Ver · 0:48

El estudio se hizo en Nepal. Se preguntó a 400 personas que usan redes sociales, entre septiembre de 2025 y enero de 2026. La mayoría tenían entre 20 y 30 años y estudios universitarios. Los que decían haber visto campañas de información o tener más educación dijeron sentirse mejor preparados contra los videos falsos. La infraestructura de ciberseguridad, las leyes y las medidas de las plataformas mostraron una relación débil o negativa con la lucha contra los videos falsos. El estudio solo mide lo que la gente percibe, no si de verdad se detectan o se frenan los videos falsos.

El estudio no nombra ninguna forma de que usted revise esto por su cuenta.

El estudio no cita ninguna autoridad electoral de Nepal ni de ningún otro país.

Qué significa para usted

En Nepal, quienes dijeron haber visto campañas de información o tener más educación se sintieron mejor preparados contra los videos falsos, según una encuesta a 400 personas que usan redes sociales. Ese estudio no muestra si de verdad se detectan o se frenan: solo recoge lo que la gente percibe. Antes de votar, fíjese en lo que su autoridad electoral publica y audita, y pregunte qué se revisó.

Kumari, A. (2026). AI and Cybersecurity: Combating Deepfakes and Misinformation in Nepal. Nepal Journal of Multidisciplinary Research. https://doi.org/10.3126/njmr.v9i2.95265

Quién pagó: El artículo no dice quién financió el estudio, si los financiadores tuvieron alguna influencia, ni quién prestó equipos o software.

Versión detalladaLos pasajes copiados del artículo, las ilustraciones y cada fuente con cuánto leímos de ella · 1224 palabras · unos 6 minLeerla →Cerrar

La regla y la evidencia: en una encuesta hecha en Nepal con 400 personas que usan medios digitales, la educación y las campañas de información aparecieron asociadas a una mayor percepción de eficacia para combatir videos falsos y noticias inventadas

### EL ACTO

No hay en el documento un acto con fuerza de ley: no hay resolución, sentencia, auditoría ni proyecto citado como vigente, con fecha de entrada en vigor ni autoridad que lo haga cumplir. Lo que el artículo registra es el resultado de una encuesta académica, no una norma, y sus recomendaciones no obligan a nadie.

Lo único que el documento pone por escrito como mandato es una recomendación dirigida a quienes deciden: "policymakers, educational institutions, cybersecurity agencies, and social media platforms prioritize large-scale digital literacy and public awareness initiatives"¹. Una recomendación no es un acto en vigor: no dice desde cuándo rige, ni a quién obliga, ni qué puede hacer un ciudadano si no se cumple.

El mismo documento añade una segunda recomendación de gobierno —"strengthen AI-related cybersecurity policies, encourage public-private collaboration for deepfake detection technologies, and improve transparency in platform-based moderation systems"²— y tampoco le fija fecha, autoridad ni plazo. Quien vote no encontrará aquí una regla que pueda invocar el día de la elección.

### EL CASO QUE RESPONDE

Esta semana no se encontró ningún caso real que se pueda contar con nombre, lugar, fecha y una segunda fuente. No se inventa ninguno.

### LO QUE SE MIDIÓ

Se preguntó a 400 personas en Nepal, de una lista de unos 500 contactadas, entre el 3 de septiembre de 2025 y el 21 de enero de 2026, cuánto creían que ayudaban cinco cosas a combatir los videos y noticias falsas³. No se midió si esas personas detectaron de verdad un video manipulado; se midió lo que percibían.

En esa medición, dos factores aparecieron asociados a una mayor percepción de eficacia: las campañas de concienciación pública (B = 0.476, p < 0.001) y la alfabetización digital (B = 0.177, p = 0.008)³. Los otros tres —infraestructura de ciberseguridad, políticas de gobierno y medidas de las plataformas— no mostraron un efecto significativo o mostraron un efecto débil³.

El conjunto de esos cinco factores explicó el 35.3% de la variación en las respuestas; el 64.7% restante quedó sin explicar⁴. En palabras llanas: la mayor parte de lo que explica por qué unas personas perciben más eficacia que otras queda fuera de este estudio.

Las medidas de las plataformas aparecieron con signo negativo (β = -0.126, p = 0.041)⁵, y los propios autores advierten que no pueden explicar del todo ese resultado porque el diseño cuantitativo no permite explorar en profundidad lo que los encuestados pensaban de esas medidas⁶.

Otras investigaciones sobre el mismo terreno no fueron encuestas de percepción. Un estudio sobre el uso de imágenes generadas por inteligencia artificial en campañas alemanas analizó partidos y encontró que los partidos minoritarios usaron esas herramientas más que los grandes, que los partidos principales declararon más a menudo el origen artificial de sus imágenes, y que la ventaja de interacción que traen esos contenidos benefició por igual a grandes y pequeños⁷⁸⁹.

Otro trabajo, en Pakistán, clasificó 17,000 noticias políticas de las elecciones generales de 2024 y halló un vínculo estadísticamente significativo entre el tono de la prensa y los resultados electorales oficiales¹⁰¹¹. Un tercero, en Nigeria, reunió a 84 personas en 12 grupos de discusión y encontró que la desinformación se volvió rutina cotidiana, que la gente desconfía del gobierno, de los medios y de las plataformas, y que casi no usa los servicios formales de verificación¹²¹³¹⁴.

En cuanto a si esto se ha medido alguna vez en una elección y no en un laboratorio o en una encuesta: el documento no lo dice. No hay en los pasajes una medición de detección real de videos falsos en una jornada electoral.

### LA RESPUESTA DE OTRA REGIÓN A LO MISMO

Esta semana no encontramos ningún tribunal, ley ni autoridad de otra región que se pronuncie sobre este método en particular.

### LA LECTURA DE LA BIBLIOTECA

Así lo leemos nosotros. Un rumor falso sobre alguien que se presenta a un cargo casi nunca inventa el mundo desde cero: toma algo que existió, lo recorta, lo saca de su momento y lo repite hasta que suena a verdad. Lo nuevo no es la mentira, es la velocidad y el parecido. Antes hacía falta un taller, una imprenta, un equipo; ahora basta un programa que aprende a imitar una cara y una voz. Por eso la pregunta útil ya no es si el video es falso —casi nunca lo parece— sino de dónde salió y quién lo puso a circular primero.

Lo que un público puede comprobar solo es poco: si el video respira raro, si la boca no cuadra con el sonido, si la fecha no cierra. Lo que no puede comprobar solo es si alguien con autoridad revisó eso y lo firmó. Y aquí conviene mirar la historia electoral de la región, que ya nos enseñó algo incómodo: las mentiras más eficaces sobre un candidato fueron las que usaron material verdadero, reciclado y fuera de contexto, no las que se inventaron todo. Eso significa que la defensa no puede ser solo "aprenda a ver", porque a simple vista el material reciclado se ve perfectamente real.

Las campañas de concienciación y la alfabetización digital aparecieron en este estudio como lo más asociado a sentirse capaz de combatir lo falso. Nosotros esperamos, en casas como la suya, que la gente confíe más en lo que puede tocar —un taller, una charla, un vecino que verifica— que en sistemas que no ve funcionar. Nos equivocaríamos si usted y sus vecinos dijeran que las leyes y las plataformas sí les resultan claras, visibles y confiables para detectar un video manipulado. Y hay un límite que debemos decir en voz alta: este resultado viene de Nepal, de 400 personas en su mayoría jóvenes y con estudios, y no se ha demostrado que se sostenga en su país, en su elección ni en su comunidad. Lo que puede hacer con esto es sencillo y concreto: antes de votar, pregunte a su autoridad electoral si tiene un sistema de verificación, quién lo audita y dónde se publica el informe. Si la respuesta es que existe pero nadie lo ha visto, ya sabe algo.

### LO QUE NO SE SABE

No se sabe si las campañas de concienciación o la alfabetización digital hacen que alguien detecte mejor un video falso, o si simplemente van juntas con otras cosas; el propio diseño del estudio no permite decirlo⁶⁴.

No se sabe por qué las medidas de las plataformas aparecieron con signo negativo: los autores escriben que ese resultado "cannot be fully explained" con este diseño⁶.

No se sabe si algo de esto se sostiene fuera de Nepal, ni en una jornada electoral real, porque el documento no lo mide.

No se sabe qué habría que verificar en su propia elección: el registro, la campaña, el conteo. El documento no dice nada de eso, y lo que no está en el registro no se escribe aquí.

Usted vota solo, en una fecha fija, y lo que ve antes de votar puede haber sido fabricado. La pregunta que le queda no es si el video es falso, sino a quién puede pedirle el informe que lo confirma.

De dónde sale cada dato de contexto, y cuánto leímos de cada documento

Kumari, A. (2026). AI and Cybersecurity: Combating Deepfakes and Misinformation in Nepal. Nepal Journal of Multidisciplinary Research. https://doi.org/10.3126/njmr.v9i2.95265 - el artículo del que trata esta nota — el artículo completo — el pasaje: "Based on these findings, the study recommends that policymakers, educational institutions, cybersecurity agencies, and social media platforms prioritize large-scale digital literacy and public awareness initiatives. Media literacy education should be integrated into academic curricula, while awareness campaigns should focus on content verification, critical thinking, and responsible online behavior."
Kumari, A. (2026). AI and Cybersecurity: Combating Deepfakes and Misinformation in Nepal. Nepal Journal of Multidisciplinary Research. https://doi.org/10.3126/njmr.v9i2.95265 - el artículo del que trata esta nota — el artículo completo — el pasaje: "Additionally, the government should strengthen AI-related cybersecurity policies, encourage public-private collaboration for deepfake detection technologies, and improve transparency in platform-based moderation systems."
Kumari, A. (2026). AI and Cybersecurity: Combating Deepfakes and Misinformation in Nepal. Nepal Journal of Multidisciplinary Research. https://doi.org/10.3126/njmr.v9i2.95265 - el artículo del que trata esta nota — el artículo completo — el pasaje: "The findings reveal that Public Awareness Campaigns and Digital Literacy significantly and positively influence the effectiveness of combating deepfakes and misinformation. Regression analysis indicates that Public Awareness Campaigns (B = 0.476, p < 0.001) and Digital Literacy (B = 0.177, p = 0.008) are significant predictors, while Cybersecurity Infrastructure, Government Policies, and Social Media Platform Measures show non-significant or weak effects."
Kumari, A. (2026). AI and Cybersecurity: Combating Deepfakes and Misinformation in Nepal. Nepal Journal of Multidisciplinary Research. https://doi.org/10.3126/njmr.v9i2.95265 - el artículo del que trata esta nota — el artículo completo — el pasaje: "This suggests that approximately 64.7% of the variation remains unexplained by the model. Therefore, other factors not examined in this study, such as trust in media, political orientation, social network influence, fact-checking behavior, technological experience, and demographic characteristics, may also contribute to individuals' ability to identify and resist misinformation."
Kumari, A. (2026). AI and Cybersecurity: Combating Deepfakes and Misinformation in Nepal. Nepal Journal of Multidisciplinary Research. https://doi.org/10.3126/njmr.v9i2.95265 - el artículo del que trata esta nota — el artículo completo — el pasaje: "Specifically, SMPM has a notable negative effect (β = -0.126, p = 0.041), implying that existing actions taken by social media platforms may be ineffective or even harmful in reducing misinformation within Nepal's context."
Kumari, A. (2026). AI and Cybersecurity: Combating Deepfakes and Misinformation in Nepal. Nepal Journal of Multidisciplinary Research. https://doi.org/10.3126/njmr.v9i2.95265 - el artículo del que trata esta nota — el artículo completo — el pasaje: "A further limitation of this study is that the quantitative design does not allow for an in-depth exploration of respondent's perceptions regarding social media platform measures. Consequently, the significant negative relationship observed between SMPM and combating deepfakes and misinformation cannot be fully explained."
Kruschinski S, Votta F. (2026). Party Equalization or Normalization Through Visual Generative AI in the 2025 German Federal Election. Media and Communication. 10.17645/mac.11859 — solo el resumen - no se pudo obtener el texto completo — el pasaje: "Regarding adoption, minor parties used VGenAI at higher rates than major parties, supporting the equalization hypothesis, which states that low-cost technologies enable resource-constrained actors to produce professional campaign visuals."
Kruschinski S, Votta F. (2026). Party Equalization or Normalization Through Visual Generative AI in the 2025 German Federal Election. Media and Communication. 10.17645/mac.11859 — solo el resumen - no se pudo obtener el texto completo — el pasaje: "Content strategy analysis reveals a transparency divide, with mainstream major parties disclosing AI origins more frequently than minor parties or the right-wing Alternative for Germany (AfD)."
Kruschinski S, Votta F. (2026). Party Equalization or Normalization Through Visual Generative AI in the 2025 German Federal Election. Media and Communication. 10.17645/mac.11859 — solo el resumen - no se pudo obtener el texto completo — el pasaje: "However, engagement analysis supports normalization: While VGenAI content is associated with higher user engagement than non-AI posts, this advantage accrues equally to major and minor parties rather than providing resource-constrained actors with competitive benefits."
Abid YA, Kashif M, Ferzund J, Hassan SR, Sheraz M, Chuah TC. (2026). Leveraging NLP and deep learning models for political opinion mining: sentiment classification and sequential analysis of news media influence on elections. PeerJ Computer Science. 10.7717/peerj-cs.4001 — solo el resumen - no se pudo obtener el texto completo — el pasaje: "A dataset of 17,000 political news stories gathered during the general elections in Pakistan in 2024 was manually classified into three categories: good, negative, and neutral."
Abid YA, Kashif M, Ferzund J, Hassan SR, Sheraz M, Chuah TC. (2026). Leveraging NLP and deep learning models for political opinion mining: sentiment classification and sequential analysis of news media influence on elections. PeerJ Computer Science. 10.7717/peerj-cs.4001 — solo el resumen - no se pudo obtener el texto completo — el pasaje: "Additionally, sentiment indices and electoral results show a statistically significant link, suggesting a quantifiable relationship between media mood and voter behavior."
Ogbodo JN, Orji-Egwu AO, Njoku EO, Ngwu OL. (2026). Dynamics and contours of online disinformation based on perspectives from Nigerian internet users. Discover Global Society. 10.1007/s44282-026-00392-9 — solo el resumen - no se pudo obtener el texto completo — el pasaje: "A qualitative design was adopted, using 12 focus group discussions conducted across Nigeria’s six geopolitical zones (urban and semi-urban locations), with 84 participants aged 18–45."
Ogbodo JN, Orji-Egwu AO, Njoku EO, Ngwu OL. (2026). Dynamics and contours of online disinformation based on perspectives from Nigerian internet users. Discover Global Society. 10.1007/s44282-026-00392-9 — solo el resumen - no se pudo obtener el texto completo — el pasaje: "Data were thematically analysed, and findings show that disinformation is routinised as an everyday feature of social media experience, generating both selective skepticism and exhaustion that erode trust in government, legacy media, and platforms."
Ogbodo JN, Orji-Egwu AO, Njoku EO, Ngwu OL. (2026). Dynamics and contours of online disinformation based on perspectives from Nigerian internet users. Discover Global Society. 10.1007/s44282-026-00392-9 — solo el resumen - no se pudo obtener el texto completo — el pasaje: "Yet, their knowledge and use of formal fact-checking organisations remain minimal."

Kumari, A. (2026). AI and Cybersecurity: Combating Deepfakes and Misinformation in Nepal. Nepal Journal of Multidisciplinary Research. https://doi.org/10.3126/njmr.v9i2.95265

Quién pagó: El artículo no dice quién financió el estudio, si los financiadores tuvieron alguna influencia, ni quién prestó equipos o software.

survey · Nepal Journal of Multidisciplinary Research · the paper, 1 Jul 2026 · free

A Nepal survey found awareness and digital skills tracked with spotting fake video, but it measured what 400 people believed, not what stopped a fake.

Short version · the longer version follows, about 6 min

Ask the weeklyAI chat

Ask me about this study: who was studied, what it found, and what it does not say.

Conversations are saved for as long as weeklyAI exists, to improve the publication. Answers come in the language you write in.

The study at a glance
Who: regular digital media users
How many: 400
Where: Nepal
When: September 3, 2025, to January 21, 2026
Kind of study: survey
Who did it: University of Wolverhampton and Herald College Kathmandu
The limit that matters: It measured what 400 people believed, not what stopped a fake.

The survey could not test whether anything actually stopped a false video.
weeklyAI's reading

Watch · 0:45

A researcher surveyed 400 regular digital media users across Nepal between September 2025 and January 2026 about deepfakes and misinformation. She asked what they thought of five responses: cybersecurity systems, digital literacy, government policy, platform moderation and public awareness campaigns. Only awareness campaigns and digital literacy were linked to a sense the fight was working. The survey could not test whether anything actually stopped a false video. The survey left about 64.7 in 100 of the differences it measured unexplained. The article says other things it did not examine, such as trust in media and political views, may also play a part.

The survey names no way for a voter to check a suspicious video or message tonight, and it names no electoral authority.

It also reports no rule, ruling or audit.

What this means for you

What this means for you as a voter is limited, and the survey itself says so. It asked 400 people in Nepal what they believed about fighting fake video, and those who followed awareness campaigns or had digital skills felt the fight was going better. That is a feeling, not proof that anything stopped a false video, so watch for audits and rulings rather than trusting a single clip.

Kumari, A. (2026). AI and Cybersecurity: Combating Deepfakes and Misinformation in Nepal. Nepal Journal of Multidisciplinary Research. https://doi.org/10.3126/njmr.v9i2.95265

Who paid: The article does not say who funded the study, whether funders had any say, or who lent equipment or software.

The longer versionThe passages copied from the paper, the pictures, and every source with how much of it we read · 1238 words · about 6 minRead it →Close

The rule and the evidence: A Nepal survey found awareness and digital skills tracked with spotting fake video, but it measured what 400 people believed, not what stopped a fake

### THE ACT

This week's document is not a law, a ruling or an audit. It is a survey study, and what stands in the place of an act is its own design. It was carried out in Nepal, its data were collected between September 3, 2025, and January 21, 2026, and it was received on April 30, 2026, revised and accepted on June 29, 2026, and published in Vol. 9, No. 2, June 2026¹.

The study set out to examine "the effectiveness of strategies used to detect and reduce deepfakes and misinformation in Nepal"². It asked 400 people across Nepal to answer a structured online questionnaire, and it measured everything on a five-point scale from strongly disagree to strongly agree². The people who took part were regular users of digital media and social networking platforms².

The study's own recommendation is what a person can act on. It recommends that policymakers, educational institutions, cybersecurity agencies and social media platforms "prioritize large-scale digital literacy and public awareness initiatives," that media literacy education be integrated into academic curricula, and that awareness campaigns focus on content verification, critical thinking and responsible online behavior³. It also recommends that the government strengthen AI-related cybersecurity policies, encourage public-private collaboration for deepfake detection technologies, and improve transparency in platform-based moderation systems⁴. Nothing in the document binds anyone, sets a deadline, or names an authority that would enforce it.

### THE CASE IT ANSWERS

No real instance was found this week.

### WHAT WAS MEASURED

What was measured was not whether any measure actually stops a false video. It was what 400 respondents in Nepal perceived about the effectiveness of efforts to identify and control deepfakes and misinformation². The survey was conducted in one country, Nepal, among people who regularly use digital media and social networking platforms².

On that measure, two things moved together with it and three did not. Public Awareness Campaigns came in with a coefficient of 0.476 and a p value below 0.001, and Digital Literacy with a coefficient of 0.177 and a p value of 0.008². Cybersecurity Infrastructure, Government Policies and Regulations, and Social Media Platform Measures showed non-significant or weak effects². One of those three, Social Media Platform Measures, came in with a negative coefficient of minus 0.126 and a p value of 0.041⁵.

The whole model explained 35.3 percent of the variation in that perceived effectiveness, which leaves about 64.7 percent unexplained⁶. The study itself says other factors it did not examine — trust in media, political orientation, social network influence, fact-checking behavior, technological experience and demographic characteristics — may also contribute to a person's ability to identify and resist misinformation⁶. The study was never run in an election, and no passage here gives any measurement taken in an election rather than in this survey.

### THE OTHER REGION'S ANSWER TO THE SAME THING

A study of German parties examined whether visual generative artificial intelligence serves as an equalizing force for minor parties or reinforces existing power asymmetries in political communication⁷. It found that minor parties used these tools at higher rates than major parties, supporting the equalization hypothesis, which states that low-cost technologies enable resource-constrained actors to produce professional campaign visuals⁸. It found a transparency divide, with mainstream major parties disclosing AI origins more frequently than minor parties or the right-wing Alternative for Germany⁹. It found that while AI content is associated with higher user engagement than non-AI posts, this advantage accrues equally to major and minor parties rather than providing resource-constrained actors with competitive benefits¹⁰. It concluded that these tools appear to be associated with broader access to professional visual production while leaving engagement asymmetries intact¹¹. This is a study of party communication, not a court ruling or a law, and it is not a rule that binds anyone.

### THE LIBRARY'S READING

Here is how we read it. A false rumor about a candidate has always worked the same way before a ballot is marked: it plants a doubt that cannot be fully cleared, and the doubt outlives the correction. What is new is not the doubt. What is new is the speed, the cheapness and the realism of the thing that plants it.

A public can learn to check a source, to look for the same story elsewhere, to ask who made a picture and when. A public cannot check what it never sees: how a feed chose what to show, what was left out, and why. That gap is where the old rumor and the new machine meet.

So when a post arrives with many likes and shares, ask whether you trust it because it is true or because others seem to trust it. When a feed shows you a video of someone saying something shocking, ask who selected what you are seeing and what might have been left out. Before you share, find the original — and if you cannot find it, that is your answer.

### WHAT IS NOT KNOWN

The study measured what 400 Nepali respondents perceived, not actual detection or prevention of deepfakes, so it cannot tell you whether any measure really stops a false video before an election².

The authors themselves say the quantitative design does not allow for an in-depth exploration of respondents' perceptions regarding social media platform measures, and that the significant negative relationship they observed between those measures and combating deepfakes and misinformation cannot be fully explained¹².

The sample was mostly young and educated: 72.8 percent were aged 20 to 30, and 52.5 percent held a bachelor's degree². The study is from Nepal and says nothing about Latin America, the United States or Canada.

The study used purposive sampling and self-reported answers on a five-point scale, and a cross-sectional design cannot show whether awareness and literacy cause better resistance or simply travel with it². Cybersecurity Infrastructure and Government Policies and Regulations were highly correlated with each other, at 0.811, which the study says may require attention in further regression analysis².

We also read that findings of this kind have not been shown to carry over to other places, populations or systems — that is about carrying over, not about this result. We also read that systems of this kind carry a known risk to watch for: when the selection of what we see is made by others for reasons we cannot inspect, our confidence can rise while our checking falls.

Here is how we read it, and what you can do with it. The pattern is simple enough: what people are taught to question, they question; what is done to them out of sight, they cannot. So expect that in homes like yours the strongest protection will not arrive as a new tool or a new rule but as someone — a teacher, a reporter, a relative — showing you how to trace a claim back to its origin.

You would know we are wrong if people who check sources carefully turned out to be fooled just as often as those who do not. Until then, here is the thing to carry to the screen: before you believe a video of anyone saying anything, find the original. If there is no original, you have your answer — and so does the person who sent it to you.

Where each piece of context comes from, and how much of it we read

Kumari, A. (2026). AI and Cybersecurity: Combating Deepfakes and Misinformation in Nepal. Nepal Journal of Multidisciplinary Research. https://doi.org/10.3126/njmr.v9i2.95265 - the article this story is about — the whole article — the passage: "Deepfake technology represents one of the most disruptive applications of AI, enabling the creation of highly realistic synthetic audio-visual content. These systems rely on deep learning models that can replicate human facial expressions, voice patterns, and behaviors with high accuracy (Mirsky & Lee, 2021)."
Kumari, A. (2026). AI and Cybersecurity: Combating Deepfakes and Misinformation in Nepal. Nepal Journal of Multidisciplinary Research. https://doi.org/10.3126/njmr.v9i2.95265 - the article this story is about — the whole article — the passage: "The findings reveal that Public Awareness Campaigns and Digital Literacy significantly and positively influence the effectiveness of combating deepfakes and misinformation. Regression analysis indicates that Public Awareness Campaigns (B = 0.476, p < 0.001) and Digital Literacy (B = 0.177, p = 0.008) are significant predictors, while Cybersecurity Infrastructure, Government Policies, and Social Media Platform Measures show non-significant or weak effects."
Kumari, A. (2026). AI and Cybersecurity: Combating Deepfakes and Misinformation in Nepal. Nepal Journal of Multidisciplinary Research. https://doi.org/10.3126/njmr.v9i2.95265 - the article this story is about — the whole article — the passage: "Based on these findings, the study recommends that policymakers, educational institutions, cybersecurity agencies, and social media platforms prioritize large-scale digital literacy and public awareness initiatives. Media literacy education should be integrated into academic curricula, while awareness campaigns should focus on content verification, critical thinking, and responsible online behavior."
Kumari, A. (2026). AI and Cybersecurity: Combating Deepfakes and Misinformation in Nepal. Nepal Journal of Multidisciplinary Research. https://doi.org/10.3126/njmr.v9i2.95265 - the article this story is about — the whole article — the passage: "Additionally, the government should strengthen AI-related cybersecurity policies, encourage public-private collaboration for deepfake detection technologies, and improve transparency in platform-based moderation systems."
Kumari, A. (2026). AI and Cybersecurity: Combating Deepfakes and Misinformation in Nepal. Nepal Journal of Multidisciplinary Research. https://doi.org/10.3126/njmr.v9i2.95265 - the article this story is about — the whole article — the passage: "Specifically, SMPM has a notable negative effect (β = -0.126, p = 0.041), implying that existing actions taken by social media platforms may be ineffective or even harmful in reducing misinformation within Nepal's context."
Kumari, A. (2026). AI and Cybersecurity: Combating Deepfakes and Misinformation in Nepal. Nepal Journal of Multidisciplinary Research. https://doi.org/10.3126/njmr.v9i2.95265 - the article this story is about — the whole article — the passage: "This suggests that approximately 64.7% of the variation remains unexplained by the model. Therefore, other factors not examined in this study, such as trust in media, political orientation, social network influence, fact-checking behavior, technological experience, and demographic characteristics, may also contribute to individuals' ability to identify and resist misinformation."
Kruschinski S, Votta F. (2026). Party Equalization or Normalization Through Visual Generative AI in the 2025 German Federal Election. Media and Communication. 10.17645/mac.11859 — only the abstract - the full text could not be fetched — the passage: "This study examines whether visual generative artificial intelligence (VGenAI) serves as an equalizing force for minor parties or reinforces existing power asymmetries in political communication."
Kruschinski S, Votta F. (2026). Party Equalization or Normalization Through Visual Generative AI in the 2025 German Federal Election. Media and Communication. 10.17645/mac.11859 — only the abstract - the full text could not be fetched — the passage: "Regarding adoption, minor parties used VGenAI at higher rates than major parties, supporting the equalization hypothesis, which states that low-cost technologies enable resource-constrained actors to produce professional campaign visuals."
Kruschinski S, Votta F. (2026). Party Equalization or Normalization Through Visual Generative AI in the 2025 German Federal Election. Media and Communication. 10.17645/mac.11859 — only the abstract - the full text could not be fetched — the passage: "Content strategy analysis reveals a transparency divide, with mainstream major parties disclosing AI origins more frequently than minor parties or the right-wing Alternative for Germany (AfD)."
Kruschinski S, Votta F. (2026). Party Equalization or Normalization Through Visual Generative AI in the 2025 German Federal Election. Media and Communication. 10.17645/mac.11859 — only the abstract - the full text could not be fetched — the passage: "However, engagement analysis supports normalization: While VGenAI content is associated with higher user engagement than non-AI posts, this advantage accrues equally to major and minor parties rather than providing resource-constrained actors with competitive benefits."
Kruschinski S, Votta F. (2026). Party Equalization or Normalization Through Visual Generative AI in the 2025 German Federal Election. Media and Communication. 10.17645/mac.11859 — only the abstract - the full text could not be fetched — the passage: "VGenAI thus appears to be associated with broader access to professional visual production while leaving engagement asymmetries intact."
Kumari, A. (2026). AI and Cybersecurity: Combating Deepfakes and Misinformation in Nepal. Nepal Journal of Multidisciplinary Research. https://doi.org/10.3126/njmr.v9i2.95265 - the article this story is about — the whole article — the passage: "A further limitation of this study is that the quantitative design does not allow for an in-depth exploration of respondent's perceptions regarding social media platform measures. Consequently, the significant negative relationship observed between SMPM and combating deepfakes and misinformation cannot be fully explained."

Kumari, A. (2026). AI and Cybersecurity: Combating Deepfakes and Misinformation in Nepal. Nepal Journal of Multidisciplinary Research. https://doi.org/10.3126/njmr.v9i2.95265

Who paid: The article does not say who funded the study, whether funders had any say, or who lent equipment or software.