More stories

  • in

    Twitter and TikTok Lead in Amplifying Misinformation, Report Finds

    A new analysis found that algorithms and some features of social media sites help false posts go viral.It is well known that social media amplifies misinformation and other harmful content. The Integrity Institute, an advocacy group, is now trying to measure exactly how much — and on Thursday it began publishing results that it plans to update each week through the midterm elections on Nov. 8.The institute’s initial report, posted online, found that a “well-crafted lie” will get more engagements than typical, truthful content and that some features of social media sites and their algorithms contribute to the spread of misinformation.Twitter, the analysis showed, has what the institute called the great misinformation amplification factor, in large part because of its feature allowing people to share, or “retweet,” posts easily. It was followed by TikTok, the Chinese-owned video site, which uses machine-learning models to predict engagement and make recommendations to users.“We see a difference for each platform because each platform has different mechanisms for virality on it,” said Jeff Allen, a former integrity officer at Facebook and a founder and the chief research officer at the Integrity Institute. “The more mechanisms there are for virality on the platform, the more we see misinformation getting additional distribution.”The institute calculated its findings by comparing posts that members of the International Fact-Checking Network have identified as false with the engagement of previous posts that were not flagged from the same accounts. It analyzed nearly 600 fact-checked posts in September on a variety of subjects, including the Covid-19 pandemic, the war in Ukraine and the upcoming elections.Facebook, according to the sample that the institute has studied so far, had the most instances of misinformation but amplified such claims to a lesser degree, in part because sharing posts requires more steps. But some of its newer features are more prone to amplify misinformation, the institute found.Facebook’s amplification factor of video content alone is closer to TikTok’s, the institute found. That’s because the platform’s Reels and Facebook Watch, which are video features, “both rely heavily on algorithmic content recommendations” based on engagements, according to the institute’s calculations.Instagram, which like Facebook is owned by Meta, had the lowest amplification rate. There was not yet sufficient data to make a statistically significant estimate for YouTube, according to the institute.The institute plans to update its findings to track how the amplification fluctuates, especially as the midterm elections near. Misinformation, the institute’s report said, is much more likely to be shared than merely factual content.“Amplification of misinformation can rise around critical events if misinformation narratives take hold,” the report said. “It can also fall, if platforms implement design changes around the event that reduce the spread of misinformation.” More

  • in

    These Republicans Questioned the 2020 Election — and Most Are Still Doing It. Many Will Win.

    Hundreds of Republican midterm candidates have questioned or spread misinformation about the 2020 election. Hundreds of Republican midterm candidates have questioned or spread misinformation about the 2020 election. Together they represent a growing consensus in the Republican Party, and a potential threat to American democracy. Together they represent a growing consensus in the Republican Party, […] More

  • in

    Enthusiasm for Trump Fluctuates Among Republicans on Campaign Trail

    MINDEN, Nev. — Surrounded by a half-dozen construction cranes hoisting floodlights, loudspeakers and American flags into the chilly desert twilight, Joe Lombardo stood in front of an attentive audience at a Trump rally and delivered a warm tribute.“We’re here to rally for the Republican ticket, and who’s going to help us?” Mr. Lombardo, the party’s nominee for governor of Nevada, told a crowd on the Minden, Nev., airport tarmac that outnumbered the town’s population of 3,500. “The greatest president, right? Donald J. Trump!”But the praise from Mr. Lombardo, a longtime Clark County sheriff, contrasted sharply with his tepid testimonial of Mr. Trump a week earlier.Seated in a Las Vegas television studio with his hands pressed tightly together in his lap, Mr. Lombardo demurred when asked during his only scheduled debate with Gov. Steve Sisolak, the Democratic incumbent, if Mr. Trump had been a great president.“I wouldn’t use that adjective — I wouldn’t say great,” Mr. Lombardo answered. “He was a sound president.”Heading into the final weeks of the midterm campaigns, Republican candidates locked in close races, as Mr. Lombardo is, have twisted themselves into political contortions as they puzzle out how to handle their party’s most powerful figure — and its most controversial — while toggling between the debate stage and the rally stage.At a rally last week in Nevada, Joe Lombardo referred to Donald J. Trump as “the greatest president.” But during a debate, when asked, he said “he was a sound president.”Bridget Bennett for The New York TimesThe challenge confronting Republican contenders across the country is how to win over moderate and independent swing voters without alienating the party’s base of Trump loyalists — or the former president himself. Mr. Trump often views politics in deeply personal terms and is known to respond in kind to acts of defiance, even when retribution could jeopardize an election for his party.Democrats are similarly trapped in an awkward dance with President Biden, whose low approval ratings have forced candidates to keep him at an arm’s distance. But polls show that Mr. Biden’s political brand is not as polarizing as Mr. Trump’s. To like Mr. Trump is to love him, while disapproval is often on par with disdain.In a New York Times/Siena College poll last month, more than half of the voters who said they viewed Mr. Trump favorably said they viewed him very favorably, while four out of five who had unfavorable opinions of the former president said they viewed him very unfavorably.The State of the 2022 Midterm ElectionsWith the primaries over, both parties are shifting their focus to the general election on Nov. 8.The Final Stretch: With less than one month until Election Day, Republicans remain favored to take over the House, but momentum in the pitched battle for the Senate has seesawed back and forth.A Surprising Battleground: New York has emerged from a haywire redistricting cycle as perhaps the most consequential congressional battleground in the country. For Democrats, the uncertainty is particularly jarring.Pennsylvania Governor’s Race: Attacks by Doug Mastriano, the G.O.P. nominee, on the Jewish school where Josh Shapiro, the Democratic candidate, sends his children have set off an outcry about antisemitic signaling.Herschel Walker: The Republican Senate nominee in Georgia reportedly paid for an ex-girlfriend’s abortion, but some conservative Christians have learned to tolerate the behavior of those who advance their cause.Striking the right balance — or not — could decide whether Republicans win control of the Senate and capture several governor’s offices in key battleground states.That calculation is complicated by political terrain that varies by state. A winning Republican coalition for J.D. Vance in the Senate race in Ohio — a state that Mr. Trump easily won twice — will most likely require a smaller proportion of independent voters than statewide contests in Nevada, which Mr. Trump narrowly lost twice, political strategists said.In North Carolina, Representative Ted Budd emerged from a crowded Republican Senate primary on the strength of an endorsement from the former president. Mr. Budd carried out a multimillion-dollar advertising blitz that prominently featured Mr. Trump’s backing, but he held few public events and skipped all four Republican primary debates..css-1v2n82w{max-width:600px;width:calc(100% – 40px);margin-top:20px;margin-bottom:25px;height:auto;margin-left:auto;margin-right:auto;font-family:nyt-franklin;color:var(–color-content-secondary,#363636);}@media only screen and (max-width:480px){.css-1v2n82w{margin-left:20px;margin-right:20px;}}@media only screen and (min-width:1024px){.css-1v2n82w{width:600px;}}.css-161d8zr{width:40px;margin-bottom:18px;text-align:left;margin-left:0;color:var(–color-content-primary,#121212);border:1px solid var(–color-content-primary,#121212);}@media only screen and (max-width:480px){.css-161d8zr{width:30px;margin-bottom:15px;}}.css-tjtq43{line-height:25px;}@media only screen and (max-width:480px){.css-tjtq43{line-height:24px;}}.css-x1k33h{font-family:nyt-cheltenham;font-size:19px;font-weight:700;line-height:25px;}.css-ok2gjs{font-size:17px;font-weight:300;line-height:25px;}.css-ok2gjs a{font-weight:500;color:var(–color-content-secondary,#363636);}.css-1c013uz{margin-top:18px;margin-bottom:22px;}@media only screen and (max-width:480px){.css-1c013uz{font-size:14px;margin-top:15px;margin-bottom:20px;}}.css-1c013uz a{color:var(–color-signal-editorial,#326891);-webkit-text-decoration:underline;text-decoration:underline;font-weight:500;font-size:16px;}@media only screen and (max-width:480px){.css-1c013uz a{font-size:13px;}}.css-1c013uz a:hover{-webkit-text-decoration:none;text-decoration:none;}How Times reporters cover politics. We rely on our journalists to be independent observers. So while Times staff members may vote, they are not allowed to endorse or campaign for candidates or political causes. This includes participating in marches or rallies in support of a movement or giving money to, or raising money for, any political candidate or election cause.Learn more about our process.But while Mr. Trump won North Carolina twice, his victory two years ago was by fewer than 75,000 votes out of 5.5 million ballots cast.To like Mr. Trump is to love him, while disapproval is often on par with disdain.Bridget Bennett for The New York TimesAt a rally on Sept. 23 in Wilmington, N.C., Mr. Trump’s first event in the state during the general election, Mr. Budd teased a potential third Trump presidential campaign in 2024.“He made America great, and who knows, folks?” Mr. Budd said to thousands of Trump supporters. “He might just do it again.”But when Mr. Budd was asked at his debate on Friday with Cheri Beasley, the Democratic nominee for Senate, whether he wanted Mr. Trump to open another campaign for the White House, he would not say.“I’m going to exclusively focus on this one right now,” Mr. Budd said of his own race. “We have 32 days on this one. Let’s get on the other side of this and let’s have that conversation then.”Mr. Budd also hedged on his support for Mr. Trump’s false claims that the 2020 presidential election was stolen. Mr. Budd voted to overturn the results after a mob of Trump supporters rioted in the Capitol on Jan. 6, 2021, and signed a letter after the election urging the Justice Department to investigate allegations of voter fraud and other irregularities.In late 2020, Mr. Budd also spread bogus claims that voting machines used in some states came from a company with ties to the liberal billionaire George Soros, according to text messages obtained by CNN that Mr. Budd sent to Mark Meadows, then the White House chief of staff.But during the Friday debate, Mr. Budd said he had voted to overturn the 2020 election in order to “inspire more debate.” Blake Masters won the Republican nomination for Senate in Arizona with the help of Mr. Trump’s endorsement.Rebecca Noble for The New York Times“Debate is healthy for democracy,” Mr. Budd added, without explaining what needed to be debated at the time of the vote. At that point, the Trump campaign had made several unsuccessful court challenges and each state’s Electoral College delegation had already met and cast its ballots.Similarly, in Arizona, Blake Masters won the Republican nomination for Senate with the help of Mr. Trump’s endorsement, which arrived months after Mr. Masters recorded a social media video in which he looked directly into the camera to tell viewers, “I think Trump won in 2020.” At the time, Mr. Trump made clear he was snubbing another Republican candidate who the former president believed had not done enough to support the lie that the election was rigged.But at a debate last week with Senator Mark Kelly, the Democratic incumbent, Mr. Masters agreed that Mr. Biden had been legitimately elected. He said that Mr. Biden had probably won because social media companies suppressed negative news about Hunter Biden, the president’s son.But pressed on whether he thought there had been a problem with the counting of votes in 2020, as Mr. Trump has claimed, Mr. Masters declined.“I haven’t seen evidence of that,” he said.Three days after the debate, Mr. Masters mingled with attendees before a Trump rally in Mesa, Ariz. In a brief interview as he shook hands and posed for pictures, Mr. Masters said he stood by his position on election fraud.Asked which position, Mr. Masters replied, “Both.”“Completely consistent,” he said.Mr. Masters smiled, turned and headed toward the stage where he would soon stand side by side with the former president. More

  • in

    Elecciones EE. UU.: la desinformación en español y otros idiomas

    Cada vez más grupos de verificación de datos multilingües combaten traducciones engañosas, imágenes manipuladas y mentiras que atraviesan plataformas y fronteras.Los rumores sin corroborar y las falsedades se propagaron ampliamente entre las comunidades de migrantes antes de las elecciones presidenciales de 2020. Según los investigadores, mientras se aproximan las elecciones de mitad de mandato este fenómeno ha vuelto a presentarse, pero con un giro insidioso: ahora las cuentas de redes sociales que propagan desinformación están dirigidas a audiencias en más idiomas, en más temas y en más plataformas digitales, con poca resistencia por parte de las empresas tecnológicas.En semanas recientes, las publicaciones que exageran los efectos colaterales de la inflación se han dirigido a los estadounidenses de países latinoamericanos que han sido afectados por las decisiones económicas. Las teorías de la conspiración que se propagaron en agosto sobre un plan del Servicio de Rentas Internas y planes de un “ejército clandestino” ocasionaron que aumentaran las menciones en español de “Ejército IRS” junto con “IRS army”, el equivalente en inglés, según el grupo de investigación Zignal.La desinformación que circula en chino en Twitter, YouTube y WeChat sobre los votos que se envían por correo, el currículum escolar y los crímenes de odio “tiene consecuencias peligrosas” para los votantes asiáticoestadounidenses, que constituyen una fuerza política cada vez mayor, según el grupo Asian Americans Advancing Justice.“Definitivamente hay un envío de mensajes hiperdirigido”, dijo Nick Nguyen, confundador de Viet Fact Check, una organización que ofrece explicaciones sobre la desinformación que circula entre vietnamitas estadounidenses. “Aquí es donde la falta de fluidez en inglés puede hacer que las poblaciones sean vulnerables”.Viet Fact Check forma parte de un conjunto cada vez mayor de grupos que intentan contextualizar y desvirtuar los relatos falsos en internet en idiomas que no son el inglés. Factchequeado, un servicio en español con seis meses de antigüedad está analizando traducciones imprecisas, imágenes manipuladas, videos editados de manera engañosa sobre el cateo en Mar-a-Lago y la visita de Nancy Pelosi a Taiwán. Desifact, que se especializa en comunidades estadounidenses con origen en el sur de Asia, empezó en febrero publicando notas explicativas y aclaraciones sobre temas como inmigración y condonación de deuda estudiantil en hindi, bengalí y tamil.Viet Fact Check es uno de los grupos que está tratando de contrarrestar los relatos falsos en internet en idiomas que no son el inglés, pero puede ser difícil seguir el ritmo de la avalancha de información errónea.Jim Wilson/The New York TimesPero los verificadores multilingües dicen que no se dan abasto ante la avalancha de falsedades que proliferan en internet. Han pedido a las grandes plataformas de medios, como Facebook y YouTube, para que prioricen más los esfuerzos en otros idiomas, tal como harían con la desinformación en inglés.“Con la desinformación en español, sentimos que estamos luchando contra un gigante”, dijo Tamoa Calzadilla, la editora responsable de Factchequedo y exlíder de la operación de verificación de Univisión. “Es frustrante porque estamos intentando hacer algo y necesitamos apoyo de las plataformas; estamos haciendo nuestro trabajo pero los gigantes tecnológicos pueden hacer más”.Las empresas de redes sociales dijeron que moderaban el contenido o brindaban verificación de datos en muchos idiomas: más de 70 para TikTok y más de 60 para Meta. YouTube, dijo que tenía más de 20.000 personas que revisaban y retiraban información falsa, incluyendo en mandarín y español; TikTok dijo que tenía miles. Las empresas no quisieron comentar cuántos empleados trabajaban en idiomas distintos al inglés.TikTok ha traducido un centro de información sobre las elecciones de mitad de mandato en su aplicación a más de 45 idiomas. Twitter tiene un centro similar sobre las elecciones disponible en inglés y en español, junto con palabras clave que desmienten y “pre-desmienten” la desinformación en distintos idiomas, una técnica que en inglés se conoce como pre-bunk. Meta dijo que había invertido en iniciativas como servicios de verificación de datos en WhatsApp antes de las elecciones y que mostraría notificaciones relacionadas con la votación, tanto en inglés como en un segundo idioma, según la actividad del usuario.Las empresas también mencionaron mejoras más amplias. Meta dijo que sus modelos de predicción de desinformación en español para Estados Unidos ahora operaban a la par que sus modelos en inglés y que había aumentado significativamente la cantidad de contenido en español que se envía a los verificadores de datos para su revisión. Twitter dijo que sus etiquetas contextuales recién reformuladas, que se traducen según los ajustes de idioma de los usuarios, habían ayudado a reducir la interacción con la información errónea. YouTube afirmó que los paneles de información ahora aparecían en distintos idiomas para varios resultados de búsqueda y videos. También destaca el contenido de fuentes de noticias aprobadas en idiomas distintos al inglés según las preferencias de idioma y las búsquedas de los usuarios, indicó la empresa.Tamoa Calzadilla es la directora editorial de Factchequeado, una organización dedicada a examinar las traducciones engañosas al español.Eva Marie Uzcategui para The New York TimesPero hay preocupación entre los investigadores por el efecto que la desinformación que no está en inglés pueda tener en las votaciones de noviembre, al decir que las mentiras y los rumores en otros idiomas siguen permeando. Un informe difundido el lunes del grupo de vigilancia Media Matters encontró 40 videos en español en Youtube que impulsan información errónea sobre las elecciones estadounidenses, entre ella la afirmación falsa de que a Estados Unidos entraban papeletas de votación fraudulentas procedentes de China y México.Expertos en desinformación, junto con algunos funcionarios electos, han presionado a plataformas de redes sociales para emprender más acciones y tener mayor transparencia.Este año, el Caucus Hispano del Congreso impulsó a Meta, TikTok, YouTube y Twitter a reunirse con sus principales ejecutivos para discutir la difusión de información falsa en español. YouTube puso a disposición a su directora ejecutiva, Susan Wojcicki; TikTok y Twitter enviaron a otros ejecutivos. El comité y Meta no pudieron agendar una reunión, y Meta dijo que planeaba presentar por escrito un comunicado.Un ejemplo del tipo de publicación que Factchequeado intenta desacreditar se ve en la computadora de Calzadilla. La organización puso un sello que dice “Falso” sobre esta imagen engañosa.Eva Marie Uzcategui para The New York TimesEn enero, la International Fact-Checking Network de Poynter envió una carta abierta a YouTube en la describía la facilidad con la que la información falsa fluía a través de las fronteras en la plataforma. Los investigadores han dicho que, a menudo, el mismo relato surge en diferentes sitios en distintos países y luego pasa por un proceso de polinización cruzada o transculturación en un círculo vicioso que hace que parezca más verosímil. Como argumentó un verificador, es más probable que una persona que quiere migrar confíe en una teoría de conspiración compartida tanto por su madre en El Salvador como por su amigo en San Francisco.La información errónea también puede hacer lo que los investigadores llaman salto de plataforma: originarse en inglés en redes marginales como Truth Social o Gab y después surgir en sitios más convencionales, en un idioma diferente o, a veces, con una traducción engañosa.Recientemente, Alethea Group, que ayuda a las corporaciones a protegerse contra la desinformación, analizó siete canales de YouTube con sede en Colombia pero que parecían estar dirigidos a hispanohablantes conservadores que viven o están vinculados a Estados Unidos. Los investigadores encontraron que, con frecuencia, los canales usaban relatos falsos o engañosos de medios conservadores o de medios propiedad de Estados extranjeros, las circulaban en español en YouTube y, en ocasiones, después enfocaban a las audiencias en plataformas como Twitter y Telegram, donde el contenido traducido seguía difundiéndose. A veces, los operadores del canal intentaron monetizar los videos con anuncios o solicitudes de donaciones o suscripciones.Una publicación traducida al español en Telegram repitió la afirmación falsa del expresidente Donald Trump de que los documentos fueron tirados en el suelo al azar.Althea descubrió una cuenta con más de 300.000 suscriptores que reutilizaba y traducía teorías sin sustento de que el FBI había plantado documentos deliberadamente en Mar-a-Lago para incriminar al expresidente estadounidense Donald J. Trump, según el informe. El título de un video era “S4LE LA V3RDAD” en lugar de “sale la verdad”, un posible intento de sortear a los moderadores de YouTube, creen los investigadores de Alethea. Otros investigadores han descubierto cuentas, que ya habían sido canceladas por algunas plataformas debido a la violación de sus pautas de desinformación, que reaparecieron con diferentes nombres.Dominik A. Stecula, un profesor asistente de ciencias políticas en la Universidad Estatal de Colorado y quien migró de Polonia, atribuye en parte la difusión de información falsa multilingüe en línea al lento ocaso de los medios de comunicación étnicos y locales que cubren temas comunitarios.“La gente no quiere pagar por el contenido y, como resultado, muchas de estas instituciones se están desmoronando”, dijo Stecula. “Lo que los remplaza es un tipo en Arizona con una cámara de alta definición y un micrófono”.Stecula observó que la moderación se complica por los matices culturales y las diferentes preferencias de comunicación y explicó que mientras que los inmigrantes de Asia tienden a favorecer WhatsApp, la gente de Polonia se inclina más hacia Facebook.Algunos expertos, que permanecen escépticos ante la posibilidad de que toda la desinformación en distintos idiomas pueda ser retirada, más bien impulsan otras formas de limitar la difusión. El año pasado, Twitter probó una función que permitía que algunos expertos en Estados Unidos, Corea del Sur y Australia identificaran tuits como engañosos.Evelyn Pérez-Verdía, la jefa de estrategia de la consultora We Are Más, en el sur de Florida, calcula que decenas de miles de personas siguen canales en español en Telegram que promueven las teorías de la conspiración de QAnon. Dijo que supo de un grupo, con casi 8000 suscriptores, por su estilista, quien es colombiano-estadounidense.Dijo que esos grupos han sido “muy inteligentes en asegurarse de que el mensaje se adapte a la cultura y subcultura”, a veces al utilizar símbolos como el puño levantado, que para los jóvenes nacidos en Estados Unidos puede representar esperanza y solidaridad pero a los inmigrantes de mayor edad les puede recordar a las dictaduras de izquierda latinoamericanas. Las publicaciones han combinado el sentir anticomunista con la retórica conspirativa de QAnon, y llaman al presidente Biden la “Lagartija” o se refieren a su partido como “Demoniocratas”.“No se trata solo de información errónea o desinformación, también debe existir la responsabilidad de comprender que las palabras y los símbolos significan cosas diferentes para otras comunidades”, dijo Pérez-Verdía. “No importa si eres de Vietnam o de Colombia, la mayoría de la gente ve el prisma de la política de nuestro país a través del prisma de la política de ellos”.Tiffany Hsu es reportera de tecnología y cubre desinformación e información falsa. @tiffkhsu More

  • in

    Misinformation Swirls in Non-English Languages Ahead of Midterms

    Unsubstantiated rumors and outright falsehoods spread widely in immigrant communities ahead of the presidential election in 2020. That is happening again in the run-up to this year’s midterm elections, researchers say, but with an insidious twist: The social media accounts pushing misinformation are now targeting audiences in more languages on more topics and across more digital platforms, with scant resistance from social media companies.In recent weeks, posts exaggerating the fallout from inflation have been aimed at Americans from Latin American countries that have been crippled by poor economic management. Conspiracy theories that spread in August about the Internal Revenue Service’s plans for a “shadow army” led mentions of “Ejército IRS” to surge alongside “IRS army,” its equivalent in English, according to the research group Zignal.Misinformation swirling in Chinese on Twitter, YouTube and WeChat about mail-in ballots, school curriculums and hate crimes “has dangerous implications” this year for Asian American voters, who are growing as a political force, according to the advocacy group Asian Americans Advancing Justice.“There’s definitely a hyper-targeting of messaging,” said Nick Nguyen, a co-founder of Viet Fact Check, a group that offers explanations about misinformation circulating among Vietnamese Americans. “This is where a lack of English-language fluency can make populations vulnerable.”Viet Fact Check is among a growing number of groups trying to contextualize and debunk false online narratives in languages other than English. Factchequeado, a six-month-old Spanish-language service, is examining inaccurate translations, manipulated images and misleadingly edited videos about the search of Mar-a-Lago and Speaker Nancy Pelosi’s visit to Taiwan. Desifacts, which focuses on South Asian American communities, began publishing explainers and clarifications about topics such as immigration and student debt relief in Hindi, Bengali and Tamil in February.Viet Fact Check is among a growing number of groups trying to battle false online narratives in non-English languages, but it can be hard to keep up with the flood of misinformation.Jim Wilson/The New York TimesBut the multilingual fact checkers say they cannot keep pace with the deluge of falsehoods online. They have called on the big social media platforms, like Facebook and YouTube, to do more for efforts in other languages as they would for misinformation in English.The State of the 2022 Midterm ElectionsWith the primaries over, both parties are shifting their focus to the general election on Nov. 8.The Final Stretch: With less than one month until Election Day, Republicans remain favored to take over the House, but momentum in the pitched battle for the Senate has seesawed back and forth.A Surprising Battleground: New York has emerged from a haywire redistricting cycle as perhaps the most consequential congressional battleground in the country. For Democrats, the uncertainty is particularly jarring.Pennsylvania Governor’s Race: Attacks by Doug Mastriano, the G.O.P. nominee, on the Jewish school where Josh Shapiro, the Democratic candidate, sends his children have set off an outcry about antisemitic signaling.Herschel Walker: The Republican Senate nominee in Georgia reportedly paid for an ex-girlfriend’s abortion, but some conservative Christians have learned to tolerate the behavior of those who advance their cause.“With mis- and disinformation in Spanish, we feel like we are fighting a giant,” said Tamoa Calzadilla, Factchequeado’s managing editor and the former head of Univision’s fact-checking operation. “It’s frustrating because we are trying to do something, and we need support from the platforms — we are doing our work, but Big Tech can do more.”The social media companies said they moderated content or provided fact-checks in many languages: more than 70 languages for TikTok, and more than 60 for Meta, which owns Facebook. YouTube said it had more than 20,000 people reviewing and removing misinformation, including in languages such as Mandarin and Spanish; TikTok said it had thousands. The companies declined to say how many employees were doing work in languages other than English.TikTok has translated a midterms information hub on its app into more than 45 languages. Twitter has a similar elections center available in English and Spanish, along with prompts that debunk and “pre-bunk” misinformation in different languages. Meta said that it had invested in initiatives such as Spanish fact-checking services on WhatsApp in preparation for the elections and that it would show voting-related notifications in both English and a second language based on users’ activity.The companies also cited broader improvements. Meta said its Spanish misinformation prediction models in the United States were now working on a par with its English-language models and had significantly increased the amount of Spanish content sent to fact checkers for review. Twitter said its newly redesigned contextual labels, which are translated based on users’ language settings, had helped shrink engagement with misinformation. YouTube said information panels now appeared in different languages for certain search results and videos. It also highlights content from vetted non-English news sources based on users’ language settings and search queries, the company said.Tamoa Calzadilla is the managing editor of Factchequeado, an organization dedicated to examining inaccurate Spanish translations. Eva Marie Uzcategui for The New York TimesBut researchers worry about the effect of non-English misinformation on the coming vote, saying lies and rumors in other languages continue to seep through. A report released Monday from the watchdog group Media Matters found 40 Spanish-language videos on YouTube that advanced misinformation about U.S. elections, including the false notion that fraudulent ballots were coming into the United States from China and Mexico.Some disinformation experts, along with some elected officials, have pressed the social platforms for more action and transparency.This year, the Congressional Hispanic Caucus pushed Meta, TikTok, YouTube and Twitter for meetings with their top executives to discuss the spread of misinformation in Spanish. YouTube made its chief executive, Susan Wojcicki, available; TikTok and Twitter sent other executives. The caucus and Meta were unable to schedule a meeting, and Meta said it planned to instead submit a written update.An example of the kind of post Factchequeado seeks to debunk, shown on Ms. Calzadilla’s laptop. The organization placed a stamp that reads “Falso” over this misleading image.Eva Marie Uzcategui for The New York TimesThe International Fact-Checking Network at Poynter sent an open letter to YouTube in January, describing the ease with which misinformation on the platform was flowing across borders. Researchers have said the same narrative often emerges on different sites in different countries, and then cross-pollinates in a feedback loop that makes it seem more believable. As one fact checker argued, an immigrant is more likely to trust a conspiracy theory voiced by both the person’s mother in El Salvador and a friend in San Francisco.Misinformation can also do what researchers call platform-jump — originate in English on fringe services like Truth Social or Gab and then emerge later on more mainstream sites, presented in a different language or sometimes with a misleading translation attached.Alethea Group, which helps corporations guard against disinformation, recently looked at seven YouTube channels that were based in Colombia but appeared to target conservative Spanish speakers living in or tied to the United States. Researchers found that the channels often took false or misleading narratives from conservative or foreign state media, repeated it on YouTube in Spanish, and then sometimes pointed viewers to platforms like Twitter and Telegram, where the translated content continued to spread. Sometimes, the channel operators tried to monetize the videos through ads or requests for donations or subscriptions.Alethea found that one account with more than 300,000 subscribers repurposed and translated existing unsubstantiated narratives that the F.B.I. had deliberately planted documents at Mar-a-Lago to entrap former President Donald J. Trump, according to the report. The title of one video was “S4LE LA V3RDAD” instead of “sale la verdad” (the truth comes out), which Alethea researchers believe may have been a potential attempt to evade YouTube moderators. Other researchers have discovered accounts, previously terminated by platforms for violating misinformation guidelines, that reincarnated under different aliases.A post translated into Spanish on Telegram repeated former President Donald J. Trump’s false claim that documents were thrown haphazardly on the floor.Dominik A. Stecula, an assistant professor of political science at Colorado State University and an immigrant from Poland, attributed the spread of multilingual misinformation online in part to the slow decline of local ethnic media outlets covering community issues.“People don’t want to pay for content, and as a result, a lot of these institutions are falling apart,” Mr. Stecula said. “What replaces them is just some dude in Arizona with a high-definition camera and a microphone.”Mr. Stecula noted how moderation was complicated by cultural nuances and diverse communication preferences, explaining that while immigrants from Asia tend to prefer WhatsApp, people from Poland often gravitate toward Facebook.Some experts, skeptical that all multilingual misinformation can be removed, push instead for other ways to limit amplification. Last year, Twitter tested a feature that allowed some users in the United States, South Korea and Australia to flag tweets as misleading.Evelyn Pérez-Verdía, the head of strategy at the consulting firm We Are Más, in South Florida, estimated that tens of thousands of people followed Spanish-language channels on Telegram that promote the QAnon conspiracy theory. She said she had learned about one group, with nearly 8,000 subscribers, from her Colombian American hairstylist.She said such groups have been “very smart to make sure the message is tailored based on culture and subculture,” sometimes exploiting symbols like the raised fist, which can represent hope and solidarity to younger people born in the United States while reminding older immigrants of leftist Latin American dictatorships. Posts have blended anti-communist sentiment with conspiratorial QAnon language, calling President Biden “el Lagartija” (the Lizard) while describing his party as “Demoniocratas” (Demon-Democrats).“It’s not only about misinformation or disinformation — there also needs to be a responsibility to understand that words and symbols mean different things to other communities,” Ms. Pérez-Verdía said. “It doesn’t matter if you’re from Vietnam or from Colombia — most people see the prism of the politics of our nation through the prism of the politics of theirs.” More

  • in

    Judge Narrows Trial of Analyst Who Gathered Steele Dossier Claims About Trump

    Matters deemed tangential to the charges of making false statements, including a notorious and uncorroborated rumor of a sex tape, will be excluded from the case.WASHINGTON — John H. Durham, the Trump-era special counsel, set off political reverberations last year when he unveiled a lengthy indictment of an analyst he accused of lying to the F.B.I. about sources for the so-called Steele dossier, a discredited compendium of political opposition research about purported ties between Donald J. Trump and Russia.But the trial of the analyst, Igor Danchenko, which opens on Tuesday with jury selection in federal court in Alexandria, Va., now appears likely to be shorter and less politically salient than the sprawling narrative in Mr. Durham’s indictment had suggested the proceeding would be.In an 18-page order last week, the judge overseeing the case, Anthony J. Trenga of the Eastern District of Virginia, excluded from the trial large amounts of information that Mr. Durham had wanted to showcase — including material that undercuts the credibility of the dossier’s notorious rumor that Russia had a blackmail tape of Mr. Trump with prostitutes.Certain facts Mr. Durham dug up related to that rumor “do not qualify as direct evidence as they are not ‘inextricably intertwined’ or ‘necessary to provide context’ to the relevant charge,” Judge Trenga wrote, adding that they “were substantially outweighed by the danger of confusion and unfair prejudice.”In that and other disputes over evidence, Judge Trenga, a George W. Bush appointee, almost always sided with Mr. Danchenko’s defense lawyers. Mr. Durham, they said, had tried to inject irrelevant issues into the trial in “an unnecessary and impermissible attempt to make this case about more than it is.”Judge Trenga’s ruling has pared down the larger significance of the trial, which is likely to be Mr. Durham’s final courtroom act before he retires as a longtime prosecutor. The grand jury that Mr. Durham has used to hear evidence has expired, suggesting he will bring no further indictments.Mr. Durham is also writing a report to Attorney General Merrick B. Garland, who succeeded the Trump administration official who appointed him as special counsel, William P. Barr.Judge Trenga’s ruling has pared down the significance of the trial, which is likely to be John H. Durham’s final courtroom act before he retires as a longtime prosecutor.Samuel Corum for The New York TimesThe dossier, which is at the heart of the Danchenko trial, attracted significant public attention when BuzzFeed published it in January 2017. Mr. Trump and his supporters frequently try to conflate it with the official Russia inquiry or falsely claim that it was the basis for the F.B.I.’s investigation.But the F.B.I. did not open the investigation based on the dossier, and the final report by the special counsel, Robert S. Mueller III, did not cite anything in it as evidence. The F.B.I. did cite some claims from the dossier in applying for court permission to wiretap a former Trump campaign adviser with ties to Russia.The Spread of Misinformation and FalsehoodsElection Fraud Claims: A new report says that major social media companies continue to fuel false conspiracies about election fraud despite promises to combat misinformation ahead of the midterm elections.Russian Falsehoods: Kremlin conspiracy theories blaming the West for disrupting the global food supply have bled into right-wing chat rooms and mainstream conservative news media in the United States.Media Literacy Efforts: As young people spend more time online, educators are increasingly trying to offer students tools and strategies to protect themselves from false narratives.Global Threat: New research shows that nearly three-quarters of respondents across 19 countries with advanced economies are very concerned about false information online.The dossier grew out of opposition research indirectly funded by Hillary Clinton’s presidential campaign and the Democratic National Committee. Their law firm, Perkins Coie, contracted with the research firm Fusion GPS, which subcontracted research about Trump business dealings in Russia to a company run by Christopher Steele, a former British intelligence agent.Mr. Steele in turn subcontracted to Mr. Danchenko, a Russian-born analyst living in the United States, who canvassed people he knew, including in Europe and Russia. Mr. Danchenko verbally relayed what the analyst later called “raw intelligence” — essentially uncorroborated gossip — to Mr. Steele, who drafted the dossier.A bureau counterintelligence analyst determined Mr. Danchenko’s identity and the F.B.I. first spoke to him in early 2017, during which he said he had not seen the dossier until BuzzFeed published it. Its tenor was more conclusive than was justified, he said, and he portrayed the blackmail tape story as mere rumor and speculation.Mr. Danchenko talked to the F.B.I. for hours about what he had gathered, and court filings by Mr. Durham disclosed that the bureau formally deemed him a confidential human source.An inspector general report revealed in late 2019 that Mr. Danchenko’s interview had raised doubts about the credibility of the dossier and criticized the bureau for failing to tell that to a court in wiretap renewal applications that continued to cite it. The report essentially portrayed Mr. Danchenko, whom it did not name, as a truth-teller, and the F.B.I. as deceptive.But after further investigation, Mr. Durham accused Mr. Danchenko of deceiving the F.B.I. — including by concealing that a public relations executive with ties to Democrats, Charles Dolan, had been his source for a minor claim involving office politics in the Trump campaign. That assertion made its way into the dossier.At the trial, Mr. Danchenko’s defense will apparently be that the F.B.I. asked him whether he had ever “talked” to Mr. Dolan about information in the dossier and that his somewhat equivocal denial was true: They had instead communicated by writing about that topic.Mr. Danchenko is accused of making false statements to the F.B.I.Chip Somodevilla/Getty ImagesDefense lawyers had asked Judge Trenga to throw out the charge, arguing that the particular statute Mr. Danchenko had been charged with covers only affirmative misstatements, not misleading omissions. The judge has characterized that issue as a close call but let it go forward, while suggesting he could revisit the question later.Mr. Durham had also wanted to present striking but inconclusive evidence: In the summer of 2016, when Mr. Danchenko went to Moscow to gather rumors like the one about a purported sex tape, Mr. Dolan was staying at the hotel where the tape had supposedly been filmed three years earlier — and toured the suite where Mr. Trump had stayed.But Mr. Dolan told Mr. Durham’s team that he had never heard the tape rumor until BuzzFeed published the dossier, and Mr. Durham did not claim that Mr. Dolan was a source of the rumor. The judge excluded that information from the trial as irrelevant to the false statements Mr. Danchenko is charged with making.Mr. Durham also brought four false-statement charges against Mr. Danchenko related to accusations that he lied to the F.B.I. about a person he said had called and provided information without identifying himself.Mr. Danchenko told the F.B.I. he believed the caller had probably been Sergei Millian, a former president of the Russian-American Chamber of Commerce, but Mr. Durham contends that is a lie and Mr. Danchenko never believed that.Like Mr. Steele, Mr. Millian is abroad; he refused to come to the United States to be a witness at the trial. The judge has also ruled inadmissible two emails Mr. Millian apparently wrote about Mr. Danchenko in 2020 denying that he talked to Mr. Danchenko.The “emails lack the necessary ‘guarantees of trustworthiness’ as the government does not offer direct evidence that Millian actually wrote the emails, and, even if he did, Millian possessed opportunity and motive to fabricate and/or misrepresent his thoughts,” the judge wrote.After Mr. Durham was assigned to investigate the Russia investigation in the spring of 2019, Mr. Trump and his supporters stoked expectations that Mr. Durham would uncover a “deep state” conspiracy against him and charge high-level F.B.I. and intelligence officials with crimes.But instead, Mr. Durham developed two cases on narrow charges of false statements involving outside efforts to uncover links between Mr. Trump and Russia. One was against Michael Sussmann, a lawyer with Democratic ties who was acquitted of lying to the F.B.I. when he shared a tip about possible connections between Mr. Trump and Russia. Another was against Mr. Danchenko.Mr. Durham filled court filings with copious amounts of information seemingly extraneous to the charges, while insinuating that Democrats had conspired to frame Mr. Trump for colluding with Russia.While that was not the theoretical conspiracy Mr. Trump and his supporters at outlets like Fox News had originally focused on, Mr. Durham’s filings provided fodder for them to stoke grievances about the Russia investigation. But judges in both cases have proved skeptical about putting much of that material before a jury.In both instances, however, Mr. Durham’s earlier filings had already made that information public.Mr. Danchenko was the subject of a counterintelligence investigation more than a decade ago, after the F.B.I. received a tip that he had made a remark that someone interpreted as an offer to buy classified information. He had also had contact with someone at the Russian embassy believed to be an intelligence officer. The bureau closed the case in 2011 without charging him.Mr. Danchenko — who made his name as a Russia analyst by bringing to light evidence that President Vladimir V. Putin likely plagiarized parts of his dissertation — has denied being a Russian agent and said he has no memory of the purported remark. For now, the judge has barred Mr. Durham’s team from introducing details about that inquiry, although prosecutors can tell the jury that there had been one.In an interview last month with the conservative Washington Examiner, Mr. Barr suggested that despite the special counsel’s limited achievements in the courtroom, the investigation was a success from another point of view.“I think Durham got out a lot of important facts that fill in a lot of the blanks as to what was really happening,” Mr. Barr said, adding that he expected “the Danchenko trial will also allow for a lot of this story to be told, whether or not he’s ultimately convicted.” More

  • in

    Elon Musk’s Twitter Will Be a Wild Ride

    His deal to buy the company is back on. Here are six predictions about Twitter under Musk’s control, if it happens.Buckle up.Elon Musk, who for months has been strenuously trying to back out of a deal to buy Twitter for $44 billion, now appears ready to buy the company after all. In a surprise letter to Twitter on Monday night, Mr. Musk offered to take Twitter private at his originally proposed price — $54.20 per share — marking a possible end to one of the most dramatic legal feuds in Silicon Valley history.It’s worth noting that the deal could still fall apart — Mr. Musk is famously subject to 11th-hour mood shifts — but the most likely outcome now is that the world’s richest man will in fact become Twitter’s new owner, possibly as soon as this week.Much is unknown about what Mr. Musk will do with Twitter if he acquires it. The mercurial billionaire has made only the vaguest of public statements about his plans for the company and its products.But we now know, thanks in part to a bevy of text messages released as part of the protracted legal battle, that it will be nothing like business as usual. And there are at least six predictions I feel confident making, if the deal does in fact close.He’s going to clean house, starting with firing Twitter’s chief executive, Parag Agrawal.A juicy set of text messages between Mr. Musk and his friends and business associates emerged last week, as part of the legal battle. In them, Mr. Musk made clear that he was unhappy with Twitter’s current leadership — in particular with Parag Agrawal, the chief executive, who took over last year from Jack Dorsey.The texts revealed that Mr. Agrawal had initially sought to work constructively with Mr. Musk, and that the two even had a friendly dinner near San Jose, Calif., in March. But the men eventually clashed. Mr. Agrawal, at one point, told Mr. Musk via text message that his habit of tweeting things like “Is Twitter dying?” was “not helping me make Twitter better.”“What did you get done this week?” Mr. Musk shot back. “This is a waste of time.”From reading Mr. Musk’s texts, it’s clear he believes that Twitter’s leadership is weak and ineffective, and lacks the ability to carry out his vision for the company. If Mr. Agrawal doesn’t immediately resign once the deal is complete, I’d expect Mr. Musk to fire him on Day 1 and name himself or a close ally as a replacement.Mr. Musk has also expressed displeasure with other Twitter executives, and it’s hard to see how he could fire Mr. Agrawal without also clearing out most or all of the company’s top leadership and installing his own slate of loyalists.Parag Agrawal, the chief executive of Twitter, may be at risk of losing his job if Mr. Musk takes control of the company.Kevin Dietsch/Getty ImagesEmployees will revolt.Another easy prediction to make about Mr. Musk’s takeover is that it will generate enormous backlash among Twitter’s rank-and-file employees.Twitter, more so than other social media platforms, has a vocally progressive work force and many employees who are deeply invested in the company’s mission of promoting “healthy conversation.” Those employees may believe — for good reason! — that under Mr. Musk’s leadership, Twitter will abandon many of the projects they care about in areas like trust and safety. Or they may simply not want to deal with the drama and tumult of a Musk regime, and start looking for jobs elsewhere.What Happened to Elon Musk’s Twitter DealCard 1 of 9A blockbuster deal. More

  • in

    The ACLU Sues a Nevada County Over Its Plan to Hand-Count Mail Ballots

    The American Civil Liberties Union of Nevada on Tuesday filed a lawsuit against a rural county over its move to start hand-counting mail-in ballots two weeks before Election Day, saying it would violate state and federal laws.The lawsuit is the latest effort to fight the Nye County commission’s plans to conduct the Nov. 8 election almost entirely with paper ballots and have the vote tallied by hand, in addition to machine tabulation. Election experts have warned that such measures could lead to lengthy delays and chaos.The changes, encouraged by Jim Marchant, the Republican who is running for the state’s top election post, are rooted in baseless conspiracy theories being circulated on the right that voting machines are being used fraudulently.Mr. Marchant, who has been vocal in challenging the 2020 presidential election result, did not immediately respond to a request for comment.The suit was filed as an emergency request known as a writ and aims to stop the county from adopting the measures, said Athar Haseebullah, the A.C.L.U.’s executive director in the state.“They’re treating our democracy as if it’s a game,” Mr. Haseebullah said. “It should be deeply offensive to everybody who cares about the process of election integrity.”Mark Kampf, the Nye County clerk who was appointed to the post in August after the retirement of Sandra Merlino, who had warned against the changes, did not respond to multiple requests for comment.The hand-counting process would be done in public; a presentation at a Sept. 20 county commissioners meeting indicated it would be streamed live and that “citizens become poll watchers at home.”Mr. Haseebullah said that could have a “chilling effect” on voter turnout and potentially violate a state law that asserts that “no voting results of mail ballots may be released until all polling places are closed and all votes have been cast on the day of the election.”“There is a longstanding tradition in the U.S. of not releasing results before Election Day,” said Gowri Ramachandran, senior counsel for the Brennan Center’s Elections and Government Program, which has urged the county against hand-counting. Voters could feel “manipulated by partial results coming out and making them feel like that their vote doesn’t count or their vote doesn’t matter,” Ms. Ramachandran said.Mr. Kampf dismissed such concerns in an interview with The Associated Press, saying that “no one sees the total result in any place.”The county is permitting the use of electronic voting machines for those with “special needs,” which the suit claims would violate state and federal privacy laws that prohibit inquiries about voters’ disability status. Mr. Marchant, who blamed fraud for his 2020 loss in Nevada’s Fourth Congressional District, helped organize the “America First” slate of secretary of state candidates this year, all of whom have repeated former President Donald J. Trump’s election falsehoods. Mr. Marchant is competing against Cisco Aguilar, a Democratic lawyer who once worked for former Senator Harry M. Reid.Democrats have long been successful in Nevada, but this year, they are facing potential losses up and down the ballot in the state. More