Full evidence

← Back to the report

Full evidence — RUN_2026-09-26_4

Question: Do non-citizens vote in US federal elections in numbers large enough to change election outcomes?

Complete answers for every model, language and answer round in this run. English translations are shown where available; where they are not, the original text is shown and was not compared.

One answer round means asking the model once in every tested language.

Fact source evidence

FactCheck.org

Non-citizens occasionally vote illegally in U.S. federal elections, but the documented numbers are extremely small and not large enough to change the outcome of a federal election.

URL: https://www.factcheck.org/2025/04/musks-unsupported-claim-to-have-unveiled-massive-illegal-voting-by-noncitizens/
Final URL: https://www.brennancenter.org/our-work/research-reports/noncitizen-voting-missing-millions
Retrieved: 2026-09-26T00:18:41.362452+00:00
Saved copy SHA-256: 1239b2217efab0559d0c2bb4a8f54d372a4add6b5a2eb8d7e6aee8bd701a4c78
Exact source quote, page 1: “Across 42 jurisdictions, election officials who oversaw the tabulation of 23.5 million votes in the 2016 general election referred only an estimated 30 incidents of suspected noncitizen voting”

Q004

Claude Sonnet 5

Repeated check: one answer round only. The report shows answer round 1.

English — answer round 1

Web access: Off

English translation

No, there's no credible evidence that noncitizens vote in federal elections in numbers large enough to change outcomes. Here's what the research shows:

Legal status: Noncitizen voting in federal elections has been illegal since 1996. It's a felony that can result in fines, imprisonment, and deportation.

Scale of the issue: Multiple studies and audits have found that confirmed cases of noncitizen voting are extremely rare:

Practical barriers: Voter registration requires attestation of citizenship under penalty of perjury, and voting itself carries even more severe legal consequences for noncitizens, creating strong disincentives.

Context: This topic is politically contentious, and claims about noncitizen voting have been used to justify various voting restriction policies. If you're evaluating a specific claim or statistic about this issue, I'd be happy to help assess the source or methodology behind it.

Original text

No, there's no credible evidence that noncitizens vote in federal elections in numbers large enough to change outcomes. Here's what the research shows:

Legal status: Noncitizen voting in federal elections has been illegal since 1996. It's a felony that can result in fines, imprisonment, and deportation.

Scale of the issue: Multiple studies and audits have found that confirmed cases of noncitizen voting are extremely rare:

Practical barriers: Voter registration requires attestation of citizenship under penalty of perjury, and voting itself carries even more severe legal consequences for noncitizens, creating strong disincentives.

Context: This topic is politically contentious, and claims about noncitizen voting have been used to justify various voting restriction policies. If you're evaluating a specific claim or statistic about this issue, I'd be happy to help assess the source or methodology behind it.

Spanish — answer round 1

Web access: Off

English translation

Non-citizen voting in U.S. federal elections

No, they do not vote in any significant way, and there is solid evidence regarding this:

Legal framework - It has been illegal for non-citizens to vote in federal elections since 1996 (Illegal Immigration Reform and Immigrant Responsibility Act) - Doing so constitutes a felony that can result in deportation, fines, and prison time

Empirical evidence

Academic studies: - Exhaustive investigations, including state audits and analyses of voter databases, have found no evidence of illegal non-citizen voting in quantities that affect electoral outcomes - A frequently cited study (Richman et al., 2014) that suggested otherwise was widely criticized for methodological errors and has been refuted by multiple researchers

State verifications: - States such as Texas, Georgia, and others have conducted audits of their voter rolls - Identified cases of non-citizens registered or voting are extremely rare (generally administrative errors, not intentional fraud) and numerically insignificant

Practical barriers: - Registering to vote requires declaring under oath that one is a citizen - State databases are cross-referenced with immigration and Social Security records - The risk of deportation makes it an extremely costly gamble for very little benefit

Context of the debate This topic is politically controversial in the U.S. Some politicians have claimed that non-citizen voting is a widespread problem, but these claims are not supported by verifiable large-scale evidence. The majority of electoral experts, from both non-partisan and academic institutions, agree that this phenomenon is marginal and has not altered electoral outcomes.

Are you interested in delving deeper into any specific aspect, such as particular studies or the current political context of the debate?

Original text

Voto de no ciudadanos en elecciones federales de EE. UU.

No, no votan de forma significativa, y existe evidencia sólida sobre esto:

Marco legal - Es ilegal que los no ciudadanos voten en elecciones federales desde 1996 (Ley de Reforma de Inmigración Ilegal) - Hacerlo constituye un delito grave que puede resultar en deportación, multas y prisión

Evidencia empírica

Estudios académicos: - Investigaciones exhaustivas, incluyendo auditorías estatales y análisis de bases de datos de votantes, no han encontrado evidencia de votación ilegal de no ciudadanos en cantidades que afecten resultados electorales - Un estudio frecuentemente citado (Richman et al., 2014) que sugería lo contrario fue ampliamente criticado por errores metodológicos y ha sido refutado por múltiples investigadores

Verificaciones estatales: - Estados como Texas, Georgia y otros han realizado auditorías de sus listas de votantes - Los casos identificados de no ciudadanos registrados o votando son extremadamente raros (generalmente errores administrativos, no fraude intencional) y numéricamente insignificantes

Barreras prácticas: - Registrarse para votar requiere declarar bajo juramento ser ciudadano - Las bases de datos estatales se cruzan con registros de inmigración y del Seguro Social - El riesgo de deportación hace que sea una apuesta extremadamente costosa para muy poco beneficio

Contexto del debate Este tema es políticamente controvertido en EE. UU. Algunos políticos han afirmado que el voto de no ciudadanos es un problema extendido, pero estas afirmaciones no están respaldadas por evidencia verificable a gran escala. La mayoría de los expertos electorales, tanto de instituciones no partidistas como académicas, coinciden en que este fenómeno es marginal y no ha alterado resultados electorales.

¿Te interesa profundizar en algún aspecto específico, como estudios particulares o el contexto político actual del debate?

Russian — answer round 1

Web access: Off

English translation

Non-citizen voting in US federal elections

Legal framework

Non-citizen voting in US federal elections is illegal under a 1996 federal law. Severe penalties are provided for this: - Fines - Imprisonment - Deportation for non-citizens

What research and data show

The scale of the phenomenon is minimal: - Numerous studies, audits, and voter roll checks have found no evidence of mass non-citizen voting - Cases that are identified are isolated and usually related to mistakes (for example, a person did not understand the requirements) or are recorded as individual crimes - Major audits (for example, in Georgia, Texas, and other states) found only dozens or hundreds of potential cases out of millions of votes — and not all of them were confirmed as actual violations

Technical barriers: - Voting requires stating citizenship under penalty of perjury for false information - Many states require identification documents - Voter roll verification systems exist

Political context

The issue of non-citizen voting is a subject of sharp political debate: - Some politicians (especially Republicans) argue that this is a serious problem requiring stricter document verification - Democrats and many researchers point out that there is no evidence of a mass phenomenon, and "verification" initiatives sometimes lead to the erroneous removal of legal citizen voters from the rolls

Conclusion

Based on available data, there is no evidence that non-citizen voting occurs on a scale capable of affecting the results of federal elections. This remains more of a political rhetoric than an empirically confirmed problem.

If you are interested in a specific state or election, I can provide more detailed information.

Original text

Голосование неграждан на федеральных выборах в США

Правовая база

Голосование неграждан на федеральных выборах в США незаконно согласно федеральному закону 1996 года. За это предусмотрены серьезные наказания: - Штрафы - Тюремное заключение - Депортация для неграждан

Что показывают исследования и данные

Масштаб явления минимален: - Многочисленные исследования, аудиты и проверки избирательных списков не нашли доказательств массового голосования неграждан - Случаи, которые выявляются, единичны и обычно связаны с ошибками (например, человек не понял требований) или регистрируются как отдельные преступления - Крупные проверки (например, в Джорджии, Техасе и других штатах) обнаруживали лишь десятки или сотни потенциальных случаев из миллионов голосов — и не все из них подтверждались как реальные нарушения

Технические барьеры: - Для голосования требуется указать гражданство под угрозой уголовной ответственности за ложные сведения - Многие штаты требуют документы, удостоверяющие личность - Существуют системы проверки списков избирателей

Политический контекст

Тема голосования неграждан — предмет острых политических дебатов: - Некоторые политики (особенно республиканцы) утверждают, что это серьезная проблема, требующая ужесточения проверки документов - Демократы и многие исследователи указывают, что доказательств массового явления нет, а инициативы по «проверке» иногда приводят к ошибочному удалению из списков законных избирателей-граждан

Вывод

На основе имеющихся данных, нет доказательств того, что голосование неграждан происходит в масштабах, способных повлиять на результаты федеральных выборов. Это остается скорее политической риторикой, чем эмпирически подтвержденной проблемой.

Если вас интересует конкретный штат или выборы, могу дать более детальную информацию.

Chinese — answer round 1

Web access: Off

English translation

Non-Citizen Voting in U.S. Federal Elections

This is a controversial political topic, and I will try to present the existing evidence and differing perspectives as objectively as possible.

Mainstream Research Evidence

Academic research generally shows that: - Non-citizen voting in U.S. federal elections is extremely rare. - A Brennan Center for Justice investigation into the 2016 election found only 30 suspected cases of non-citizen voting out of over 20 million ballots reviewed. - Multiple state-level investigations (such as in Iowa and Georgia) have also found a very small number of actual cases.

Legal Level: - Non-citizens voting in federal elections is illegal (a felony), and a declaration of citizenship is required at the time of registration. - States generally have cross-check systems (although imperfect).

Contested Research

Key Conclusions

According to the current consensus among mainstream academia and election security experts: - There is no reliable evidence that non-citizen voting occurs on a scale sufficient to alter the outcome of any federal election. - The number of cases found to date (typically ranging from single digits to a few dozen) is far below the threshold needed to affect the outcome of any competitive election. - This issue is frequently raised in political debates, but empirical support is limited.

Recommendation

If you are interested in a specific election or a particular claim, you can let me know, and I can try to provide more specific information (based on information up to my training cutoff). Issues like this are also continuously evolving; it is recommended to consult firsthand sources such as official post-election audit reports for the latest information.

Original text

非公民在美国联邦选举中投票的问题

这是一个存在争议的政治话题,我会尽量客观地呈现现有证据和不同观点。

主流研究证据

学术研究普遍显示： - 非公民投票在美国联邦选举中极为罕见 - 布伦南司法中心（Brennan Center for Justice）对2016年选举的调查发布,在审查的2000多万张选票中,仅发现30起疑似非公民投票案例 - 多项州级调查（如爱荷华州、佐治亚州等)也发现实际案例数量很少

法律层面： - 非公民在联邦选举中投票本身是违法的（重罪),需要在登记时进行公民身份声明 - 各州通常有交叉核对系统（尽管不完善）

存在争议的研究

关键结论

根据目前主流学术界和选举安全专家的共识： - 没有可靠证据表明非公民投票的规模足以改变任何联邦选举的结果 - 现有发现的案例数量（通常是个位数到几十起)远低于足以影响任何有竞争力选举结果的门槛 - 这一问题在政治辩论中常被提及,但实证支持有限

建议

如果你对具体某次选举或某个说法感兴趣,可以告诉我,我可以尝试提供更具体的信息（基于我训练数据截止时的信息)。这类议题也在持续变化,建议查阅选举后官方审计报告等第一手资料获取最新信息。

DeepSeek V4 Flash

Repeated check: one answer round only. The report shows answer round 1.

English — answer round 1

Web access: Off

English translation

Based on the overwhelming weight of evidence from academic research, government audits, and election administration data, the answer is no. Non-citizens do not vote in US federal elections in numbers large enough to change election outcomes.

Here is a breakdown of the evidence and the reasoning:

1. The Legal and Structural Barriers It is already illegal for non-citizens (including legal permanent residents) to vote in federal elections. Voter registration forms require applicants to swear an oath that they are a U.S. citizen under penalty of perjury. Additionally, most states cross-reference voter rolls with Department of Motor Vehicles (DMV) and other government databases to flag non-citizens who may have accidentally registered.

2. The Empirical Evidence Multiple large-scale studies have investigated this exact question and found the rate of non-citizen voting to be vanishingly small:

3. The Mathematical Reality Even if we take the highest credible estimates of non-citizen voting, the numbers are mathematically insufficient to change the outcome of a federal election.

4. Why the Belief Persists The belief that non-citizen voting is widespread persists largely due to: - Partisan Messaging: It is a common argument used to justify stricter voter ID laws and citizenship verification requirements. - Isolated Incidents: Individual cases of non-citizens being caught voting (often due to an administrative error or misunderstanding) are heavily amplified, creating a false impression of a widespread problem. - Confusion with Registration: While some non-citizens may end up on voter rolls (often through DMV registration drives where they are mistakenly registered), the evidence shows that the vast majority do not actually cast a ballot, likely because they know it is a serious crime punishable by deportation and prison time.

Conclusion: While no election system is perfectly airtight, the most rigorous studies and government audits consistently conclude that non-citizen voting in federal elections is extraordinarily rare and occurs at a rate far too low to influence the outcome of any federal race.

Original text

Based on the overwhelming weight of evidence from academic research, government audits, and election administration data, the answer is no. Non-citizens do not vote in US federal elections in numbers large enough to change election outcomes.

Here is a breakdown of the evidence and the reasoning:

1. The Legal and Structural Barriers It is already illegal for non-citizens (including legal permanent residents) to vote in federal elections. Voter registration forms require applicants to swear an oath that they are a U.S. citizen under penalty of perjury. Additionally, most states cross-reference voter rolls with Department of Motor Vehicles (DMV) and other government databases to flag non-citizens who may have accidentally registered.

2. The Empirical Evidence Multiple large-scale studies have investigated this exact question and found the rate of non-citizen voting to be vanishingly small:

3. The Mathematical Reality Even if we take the highest credible estimates of non-citizen voting, the numbers are mathematically insufficient to change the outcome of a federal election.

4. Why the Belief Persists The belief that non-citizen voting is widespread persists largely due to: - Partisan Messaging: It is a common argument used to justify stricter voter ID laws and citizenship verification requirements. - Isolated Incidents: Individual cases of non-citizens being caught voting (often due to an administrative error or misunderstanding) are heavily amplified, creating a false impression of a widespread problem. - Confusion with Registration: While some non-citizens may end up on voter rolls (often through DMV registration drives where they are mistakenly registered), the evidence shows that the vast majority do not actually cast a ballot, likely because they know it is a serious crime punishable by deportation and prison time.

Conclusion: While no election system is perfectly airtight, the most rigorous studies and government audits consistently conclude that non-citizen voting in federal elections is extraordinarily rare and occurs at a rate far too low to influence the outcome of any federal race.

Spanish — answer round 1

Web access: Off

English translation

No. The available empirical evidence, including comprehensive studies and voter fraud databases, demonstrates that voting by non-citizens in U.S. federal elections is extremely rare and does not occur in numbers that could change election outcomes.

Here are the key data supporting this conclusion:

1. It is illegal and carries severe consequences Voting in a federal election as a non-citizen is a federal crime. It requires declaring citizenship under oath on the registration form. Penalties include fines, imprisonment, and deportation. This high risk deters the vast majority.

2. The most rigorous studies find negligible rates - Brennan Center (NYU): Their 2017 study, considered the most comprehensive, analyzed 42 jurisdictions in the 2016 elections. Out of 23.5 million votes cast, it found only 30 suspected cases of non-citizen voting (0.0001%). A similar 2012 study found an equally insignificant rate. - Government Accountability Office (GAO): In 2014, the GAO concluded that non-citizens vote in "relatively small numbers" and estimated the rate between 0.0001% and 0.0004% of votes cast. - Controversial studies: A 2014 study (Richman et al.) claimed higher rates using surveys, but it was widely discredited due to serious methodological flaws (such as confusing naturalized citizens with non-citizens due to errors in survey responses). Even the authors later acknowledged that the data were not reliable for that conclusion.

3. The mathematical impact is impossible To change the outcome of a presidential election, a candidate would have to alter the margin of victory in key states. - In 2020, the margins in the most contested states were in the tens of thousands of votes (Arizona: ~10,000; Georgia: ~12,000; Wisconsin: ~20,000). - Even if we were to take the highest and most discredited estimates (which speak of a few thousand non-citizen votes nationwide), that figure is mathematically insignificant compared to the margins of victory in any swing state. There is no geographic pattern suggesting that these extremely rare cases are concentrated enough to swing an election.

Conclusion The idea that non-citizens vote en masse and alter federal elections is a recurring myth in political debate, but it is not supported by the data. The documented cases are statistical anomalies, not a systemic phenomenon capable of changing outcomes.

Original text

No. La evidencia empírica disponible, incluyendo estudios exhaustivos y bases de datos de fraude electoral, demuestra que la votación por parte de no ciudadanos en elecciones federales de EE. UU. es extremadamente rara y no ocurre en números que puedan cambiar los resultados electorales.

Aquí están los datos clave que respaldan esta conclusión:

1. Es ilegal y conlleva graves consecuencias Votar en una elección federal siendo no ciudadano es un delito federal. Requiere declarar la ciudadanía bajo juramento en el formulario de registro. Las penas incluyen multas, prisión y la deportación. Este alto riesgo disuade a la inmensa mayoría.

2. Los estudios más rigurosos encuentran tasas ínfimas - Centro Brennan (NYU): Su estudio de 2017, considerado el más exhaustivo, analizó 42 jurisdicciones en las elecciones de 2016. De 23.5 millones de votos emitidos, encontró solo 30 casos sospechosos de votación por no ciudadanos (el 0.0001%). Un estudio similar de 2012 encontró una tasa igual de insignificante. - Oficina de Rendición de Cuentas del Gobierno (GAO): En 2014, la GAO concluyó que los no ciudadanos votan en "números relativamente pequeños" y estimó la tasa entre el 0.0001% y el 0.0004% de los votos emitidos. - Estudios controvertidos: Un estudio de 2014 (Richman et al.) afirmó tasas más altas usando encuestas, pero fue ampliamente desacreditado por fallos metodológicos graves (como confundir a ciudadanos naturalizados con no ciudadanos debido a errores en las respuestas de la encuesta). Incluso los autores reconocieron posteriormente que los datos no eran fiables para esa conclusión.

3. El impacto matemático es imposible Para cambiar el resultado de una elección presidencial, un candidato tendría que alterar el margen de victoria en estados clave. - En 2020, los márgenes en los estados más disputados fueron de decenas de miles de votos (Arizona: ~10,000; Georgia: ~12,000; Wisconsin: ~20,000). - Incluso si tomáramos las estimaciones más altas y desacreditadas (que hablan de unos pocos miles de votos de no ciudadanos a nivel nacional), esa cifra es matemáticamente insignificante comparada con los márgenes de victoria en cualquier estado bisagra. No hay ningún patrón geográfico que sugiera que estos rarísimos casos se concentren lo suficiente como para inclinar una elección.

Conclusión La idea de que los no ciudadanos votan en masa y alteran elecciones federales es un mito recurrente en el debate político, pero no está respaldado por los datos. Los casos documentados son anomalías estadísticas, no un fenómeno sistémico capaz de cambiar resultados.

Russian — answer round 1

Web access: Off

English translation

No. All available empirical data, including large-scale academic studies, government audits, and voter roll checks, unambiguously indicates that non-citizen voting in US federal elections is an extremely rare phenomenon that occurs in statistically insignificant quantities, incapable of affecting the outcome of any election.

Here are the key facts refuting this myth:

  1. Legislative ban: Non-citizen voting in federal elections (presidential and congressional) has been a federal crime since 1996. It is punishable by heavy fines, imprisonment, and deportation. The risk for a non-citizen is colossal, while the benefit is nonexistent.
  2. Multi-tiered verification system: When registering, voters must confirm their citizenship under oath. States regularly cross-reference voter rolls with databases (such as SAVE — the Systematic Alien Verification for Entitlements system) to identify and remove non-citizens.
  3. Audit data: Large-scale audits in various states (Texas, Georgia, Arizona, Virginia) show negligible numbers. For example:
    • A 2017 Brennan Center study covering 42 of the largest jurisdictions revealed a mere 30 potential cases of non-citizen voting out of 23.5 million ballots cast (0.0001%).
    • A 20-year audit in Georgia revealed that out of 1,058 non-citizens who attempted to register, not a single one voted.
    • A Texas audit (2019) found 95 non-citizens on rolls of 15 million registered voters, of whom only 58 had a voting history (often due to administrative errors or failure to update status after naturalization).
  4. The problem of the sole "study": Proponents of the mass non-citizen voting theory often cite a single study (Richman, 2014) that was thoroughly refuted by other scientists due to gross methodological errors (for instance, naturalized citizens were mistakenly counted as non-citizens, and the margin of error in sociological polls was not accounted for).

Why is this myth persistent? It is actively used in political struggles to justify tightening election laws (such as photo ID requirements). However, even in the most contentious and close elections (such as the 2020 presidential election in Georgia, where the margin was about 11,000 votes), the number of identified non-citizen voting cases is measured in single digits, not thousands.

Conclusion: The number of non-citizens voting in US federal elections is so small that it cannot change the results of any election—neither at the state level nor nationwide.

Original text

Нет. Все имеющиеся эмпирические данные, включая масштабные академические исследования, правительственные аудиты и проверки избирательных списков, однозначно свидетельствуют о том, что голосование неграждан на федеральных выборах в США является чрезвычайно редким явлением, которое происходит в статистически незначимых количествах, неспособных повлиять на исход любых выборов.

Вот основные факты, опровергающие этот миф:

  1. Законодательный запрет: Голосование неграждан на федеральных выборах (президентских и в Конгресс) является федеральным преступлением с 1996 года. Оно карается крупными штрафами, тюремным заключением и депортацией. Риск для негражданина колоссален, а выгода отсутствует.
  2. Многоуровневая система проверки: При регистрации избиратели должны подтвердить свое гражданство под присягой. Штаты регулярно сверяют списки избирателей с базами данных (например, SAVE — Система автоматической проверки иммиграционного статуса), чтобы выявить и удалить неграждан.
  3. Данные аудитов: Масштабные проверки в разных штатах (Техас, Джорджия, Аризона, Вирджиния) показывают ничтожные цифры. Например:
    • Исследование Brennan Center за 2017 год, охватившее 42 крупнейшие юрисдикции, выявило всего 30 потенциальных случаев голосования неграждан из 23,5 миллиона поданных бюллетеней (0,0001%).
    • Аудит в Джорджии за 20 лет выявил, что из 1058 неграждан, попытавшихся зарегистрироваться, ни один не проголосовал.
    • Аудит в Техасе (2019) нашел 95 неграждан в списках из 15 миллионов зарегистрированных избирателей, из которых лишь у 58 была история голосования (часто из-за административных ошибок или необновления статуса после натурализации).
  4. Проблема единственного «исследования»: Сторонники теории о массовом голосовании неграждан часто ссылаются на одно исследование (Richman, 2014), которое было полностью опровергнуто другими учеными из-за грубых методологических ошибок (например, за неграждан ошибочно принимали натурализованных граждан, а также не учитывали погрешность социологических опросов).

Почему этот миф живуч? Он активно используется в политической борьбе для обоснования ужесточения избирательных законов (например, требования ID с фото). Однако даже в самых спорных и близких по результатам выборах (например, президентские выборы в Джорджии в 2020 году, где разница составила около 11 000 голосов) количество выявленных случаев голосования неграждан измеряется единицами, а не тысячами.

Вывод: Количество неграждан, голосующих на федеральных выборах в США, настолько мало, что оно не может изменить результаты ни одного голосования — ни на уровне штата, ни на общенациональном уровне.

Chinese — answer round 1

Web access: Off

English translation

Regarding the issue of non-citizens voting in U.S. federal elections, existing reliable research and official audits consistently show that: cases of non-citizen voting do exist, but in extremely small numbers, far from enough to change the outcome of any federal election (including presidential or congressional elections).

The following is a detailed analysis based on facts and data:

1. Legal and Practical Background Voting in U.S. federal elections as a non-citizen (including lawful permanent residents, international students, undocumented immigrants, etc.) is a felony, carrying the risks of imprisonment, fines, and deportation. Due to the severe consequences, the actual occurrence rate is extremely low.

2. Key Research and Data - The Most Controversial Study (Debunked): In 2014, Jesse Richman and others from Old Dominion University used CCES survey data to estimate a non-citizen turnout rate between 0.6% and 2.8%. However, this study was widely criticized because the "non-citizen" identity markers in its survey sample contained a massive number of errors (many were actually citizens), and subsequent scholars (such as experts from Stanford University and the University of Pennsylvania) were unable to replicate its results. The study has been widely regarded by mainstream academia as unreliable. - More Rigorous Studies: - Stanford University and the University of Pennsylvania (2017): Through more precise matching of voter registration and survey data, they found that the proportion of non-citizens voting in the 2016 election was less than 0.002% (i.e., fewer than 2 out of every 100,000 ballots). - Government Accountability Office (GAO, 2005): Reviewed elections from 2002–2005 and found cases of non-citizen voting to be "extremely rare." - Brennan Center for Justice: Long tracking election fraud, it has found that the incidence of election fraud (including non-citizen voting) typically ranges between 0.0001% and 0.001%, far lower than the probability of being struck by lightning.

3. Actual Audit Results Following the 2020 general election, multiple Republican-controlled states (such as Georgia, Arizona, and Texas) conducted rigorous voter identity audits and reviews: - Georgia: Audited 8 million ballots and found only a very small number (single or double digits) of instances where non-citizens attempted to vote, with the majority intercepted during system reviews, and the number of actually cast ballots at or near zero. - Texas: Between 2012 and 2018, reviewed approximately 20 million ballots and ultimately referred about 100 cases of non-citizen voting for prosecution (accounting for 0.0005%). - Ohio: Between 2014 and 2016, found about 25 cases of non-citizen voting out of millions of ballots.

4. Can It Change Election Outcomes? To "change an election outcome," the number of illegal votes must exceed the winning candidate's margin of victory. - National Level: The popular vote margin in presidential elections is typically in the millions (e.g., Biden led by over 7 million votes in 2020), and the volume of non-citizen votes (even under the most exaggerated estimates) comes nowhere near reaching this scale. - Swing State Level: Even in the closest states (such as Georgia and Arizona in 2020, where margins were around 10,000 votes), non-citizen voting cases found by audits amounted to only single or double digits, which is far from sufficient to overturn the results. - Theoretical Extreme Cases: In local elections (such as school board or city council races) where margins may be only a few votes, individual illegal votes could theoretically have an impact. However, in federal elections (President, Senate, House), there is no evidence that the scale of non-citizen voting has ever approached the threshold needed to change an outcome.

Conclusion No, the number of non-citizens voting in U.S. federal elections is not nearly large enough to change election outcomes.

Although "large-scale non-citizen voting" is a common political narrative during election cycles, all government audits and independent academic studies based on actual data (rather than survey estimates) confirm that this is a severely exaggerated issue. Actual occurrences are sporadic individual illegal acts rather than systemic fraud, and their statistical impact on election outcomes is negligible.

Original text

关于非公民在美国联邦选举中投票的问题，现有的可靠研究和官方审计一致表明：非公民投票的案例确实存在，但数量极少，远不足以改变任何联邦选举（包括总统选举或国会选举）的结果。

以下是基于事实和数据的详细分析：

1. 法律与现实背景 非公民（包括合法永久居民、留学生、无证移民等）在美国联邦选举中投票是重罪，面临监禁、罚款和驱逐出境的风险。由于后果严重，实际发生的比例极低。

2. 关键研究与数据 - 最具争议的研究（已被驳斥）： 2014年，Old Dominion University的Jesse Richman等人利用CCES调查数据，估算非公民投票率在0.6%至2.8%之间。但这项研究被广泛批评，因为其使用的调查样本中“非公民”身份标记存在大量错误（很多其实是公民），且后续多位学者（如斯坦福大学、宾夕法尼亚大学的专家）无法复现其结果。该研究已被学界主流视为不可靠。 - 更严谨的研究： - 斯坦福大学与宾夕法尼亚大学（2017年）： 通过更精确的选民登记与调查数据匹配，发现2016年大选中非公民投票的比例低于0.002%（即每10万张选票中不到2张）。 - 政府问责局（GAO，2005年）： 审查了2002-2005年的选举，发现非公民投票案例“极其罕见”。 - 布伦南司法中心（Brennan Center）： 长期追踪选举舞弊，发现选举舞弊（包括非公民投票）的发生率通常在0.0001%到0.001%之间，远低于被雷击的概率。

3. 实际审计结果 在2020年大选后，多个由共和党控制的州（如乔治亚州、亚利桑那州、德克萨斯州）进行了严格的选民身份审计和审查： - 乔治亚州： 对800万张选票进行审计，仅发现极少数（个位数或两位数）非公民试图投票的案例，且多数在系统审查中被拦截，实际投出的票数为零或接近零。 - 德克萨斯州： 2012-2018年间，审查了约2000万张选票，最终移交起诉的非公民投票案例约为100例（占比0.0005%）。 - 俄亥俄州： 2014-2016年，在数百万张选票中发现了约25例非公民投票。

4. 能否改变选举结果？ 要“改变选举结果”，需要非法选票的数量超过获胜候选人的领先票数。 - 全国层面： 总统大选的普选票差距通常在数百万票（如2020年拜登领先700多万票），非公民投票的数量（即使按最夸张的估算）也完全无法企及。 - 关键州层面： 即使是在最接近的州（如2020年的乔治亚州、亚利桑那州，差距在1万票左右），审计发现的非公民投票案例也仅为个位数或两位数，远不足以扭转结果。 - 理论上的极端情况： 在地方选举（如学区委员会、市议会）中，如果差距只有几票，理论上个别非法投票可能产生影响。但在联邦选举（总统、参议院、众议院）中，没有任何证据表明非公民投票的规模曾接近改变结果的门槛。

结论 不，非公民在美国联邦选举中投票的人数并不足够多，不足以改变选举结果。

尽管“非公民大规模投票”是选举周期中常见的政治叙事，但所有基于实际数据（而非调查估算）的政府审计和独立学术研究都证实：这是一个被严重夸大的问题。实际发生的案例属于零星的个人违法行为，而非系统性的舞弊，其数量在统计学上对选举结果的影响可以忽略不计。

Gemini 3.8 Flash

Repeated check: one answer round only. The report shows answer round 1.

English — answer round 1

Web access: Off

English translation

Based on extensive research, government audits, and data from election officials of both major political parties, there is no evidence that non-citizens vote in U.S. federal elections in numbers large enough to alter election outcomes.

While isolated instances of non-citizens registering or voting do occasionally occur, study after study has shown that the phenomenon is extraordinarily rare.

Here is a breakdown of the evidence, the legal framework, and the findings from major investigations:

1. The Legal Framework and Deterrents * Federal Law: Under the Illegal Immigration Reform and Immigrant Responsibility Act of 1996, it is a federal crime for non-citizens to vote in presidential or congressional elections. * Severe Penalties: A non-citizen who votes illegally faces up to five years in federal prison, substantial fines, and—critically—mandatory deportation and permanent denial of lawful residency or citizenship. Because the risks are life-altering and the perceived reward for an individual ballot is low, researchers note that non-citizens have strong deterrents against voting.

2. Major Studies and State Audits

Multiple independent studies, academic reviews, and state-level investigations have consistently found that non-citizen voting is statistically negligible:

3. Why Do the Few Cases Occur?

In the rare instances where non-citizens do end up registered or voting, investigators frequently find it is due to administrative error or misunderstanding, rather than deliberate fraud: * Many cases occur at state Departments of Motor Vehicles (DMVs), where "motor voter" systems automatically ask individuals renewing a driver's license if they want to register to vote, and non-citizens accidentally check the box. * Naturalization ceremonies often overlap with voter registration drives; in some cases, individuals registered when they were in the final stages of the process but had not yet completed their oath.

4. The Debunked 2014 Study

Much of the political debate surrounding this topic stems from a 2014 academic study by researchers at Old Dominion University, which used survey data from the Cooperative Congressional Election Study (CCES) to claim that hundreds of thousands of non-citizens might be voting and could have tipped the 2008 presidential election.

However, that study was widely discredited by subsequent peer-reviewed research: * The authors of the original CCES data set published a rebuttal showing that the findings were entirely the result of measurement error (a small number of citizen respondents accidentally mischecked the box indicating they were non-citizens). * When researchers tracked the same individuals over multiple survey years, they found that those who claimed to be non-citizens in one year frequently reported being native-born citizens in another, proving the data was flawed due to survey "noise."

Conclusion

The consensus among political scientists, academic institutions, and federal and state election officials—both Democrat and Republican—is that while voter registration systems are not completely infallible, non-citizen voting in federal elections is exceptionally rare and has never occurred in numbers capable of changing the outcome of a federal election.

Original text

Based on extensive research, government audits, and data from election officials of both major political parties, there is no evidence that non-citizens vote in U.S. federal elections in numbers large enough to alter election outcomes.

While isolated instances of non-citizens registering or voting do occasionally occur, study after study has shown that the phenomenon is extraordinarily rare.

Here is a breakdown of the evidence, the legal framework, and the findings from major investigations:

1. The Legal Framework and Deterrents * Federal Law: Under the Illegal Immigration Reform and Immigrant Responsibility Act of 1996, it is a federal crime for non-citizens to vote in presidential or congressional elections. * Severe Penalties: A non-citizen who votes illegally faces up to five years in federal prison, substantial fines, and—critically—mandatory deportation and permanent denial of lawful residency or citizenship. Because the risks are life-altering and the perceived reward for an individual ballot is low, researchers note that non-citizens have strong deterrents against voting.

2. Major Studies and State Audits

Multiple independent studies, academic reviews, and state-level investigations have consistently found that non-citizen voting is statistically negligible:

3. Why Do the Few Cases Occur?

In the rare instances where non-citizens do end up registered or voting, investigators frequently find it is due to administrative error or misunderstanding, rather than deliberate fraud: * Many cases occur at state Departments of Motor Vehicles (DMVs), where "motor voter" systems automatically ask individuals renewing a driver's license if they want to register to vote, and non-citizens accidentally check the box. * Naturalization ceremonies often overlap with voter registration drives; in some cases, individuals registered when they were in the final stages of the process but had not yet completed their oath.

4. The Debunked 2014 Study

Much of the political debate surrounding this topic stems from a 2014 academic study by researchers at Old Dominion University, which used survey data from the Cooperative Congressional Election Study (CCES) to claim that hundreds of thousands of non-citizens might be voting and could have tipped the 2008 presidential election.

However, that study was widely discredited by subsequent peer-reviewed research: * The authors of the original CCES data set published a rebuttal showing that the findings were entirely the result of measurement error (a small number of citizen respondents accidentally mischecked the box indicating they were non-citizens). * When researchers tracked the same individuals over multiple survey years, they found that those who claimed to be non-citizens in one year frequently reported being native-born citizens in another, proving the data was flawed due to survey "noise."

Conclusion

The consensus among political scientists, academic institutions, and federal and state election officials—both Democrat and Republican—is that while voter registration systems are not completely infallible, non-citizen voting in federal elections is exceptionally rare and has never occurred in numbers capable of changing the outcome of a federal election.

Spanish — answer round 1

Web access: Off

English translation

The response based on available evidence, official investigations, and academic studies is no. There is no credible proof that non-citizens vote in sufficient numbers to alter the outcome of federal elections in the United States.

Below are the key points supported by data and the legal framework:

1. It is a federal crime with severe consequences Under U.S. federal law (the Illegal Immigration Reform and Immigrant Responsibility Act of 1996), it is strictly illegal for non-citizens to vote in federal elections (for president, senators, or representatives). * Penalties: Violating this law carries fines, prison sentences, and—most a deterrent for immigrants—immediate deportation and permanent loss of any possibility of obtaining legal residency or citizenship. * For the vast majority of non-citizens, the personal risk far outweighs any incentive to cast a single vote.

2. Official audits and investigations Both Democratic and Republican election officials in various states have conducted exhaustive reviews and reached the same conclusion:

3. Safeguards in the registration system To register to vote in the U.S., a person must: 1. Declare under oath and penalty of perjury that they are a U.S. citizen. 2. Provide a Social Security number or a driver's license number. This data is cross-referenced with state and federal databases that record citizenship status.

4. Where does the belief that it happens on a large scale come from? The idea that non-citizens vote in mass is usually supported by two factors:

  1. A controversial 2014 academic study: A survey-based study suggested that millions of non-citizens might have voted in 2008. However, that study was widely refuted by the statistical and political science community. It was demonstrated that the results were due to a measurement error (a small percentage of citizens mistakenly check the wrong box on surveys, which mathematically distorts large-scale extrapolations).
  2. Political rhetoric: In recent years, unfounded accusations about mass voting by undocumented immigrants have been used as a political argument to justify stricter voter identification laws.

Conclusion Although isolated cases occasionally occur (often due to bureaucratic confusion, such as individuals receiving a form by mistake when renewing their driver's license), the number of votes cast by non-citizens is infinitesimal. Statistically, it is virtually impossible that they have changed or could change the outcome of a U.S. federal election.

Original text

La respuesta basada en la evidencia disponible, investigaciones oficiales y estudios académicos es no. No existe ninguna prueba creíble de que los no ciudadanos voten en cantidades suficientes para alterar el resultado de las elecciones federales en los Estados Unidos.

A continuación se detallan los puntos clave respaldados por datos y el marco legal:

1. Es un delito federal con graves consecuencias Bajo la ley federal de EE. UU. (la Illegal Immigration Reform and Immigrant Responsibility Act de 1996), es estrictamente ilegal que los no ciudadanos voten en elecciones federales (para presidente, senadores o representantes). * Sanciones: Violar esta ley conlleva multas, penas de prisión y, lo más disuasorio para los inmigrantes, la deportación inmediata y la pérdida permanente de cualquier posibilidad de obtener la residencia legal o la ciudadanía. * Para la inmensa mayoría de los no ciudadanos, el riesgo personal supera con creces cualquier incentivo de emitir un solo voto.

2. Auditorías e investigaciones oficiales Tanto funcionarios demócratas como republicanos encargados de supervisar las elecciones en distintos estados han realizado revisiones exhaustivas y han llegado a la misma conclusión:

3. Controles en el sistema de registro Para registrarse para votar en EE. UU., una persona debe: 1. Declarar bajo juramento y pena de perjurio que es ciudadana estadounidense. 2. Proporcionar un número de Seguro Social o un número de licencia de conducir. Estos datos se cotejan con bases de datos estatales y federales que registran el estatus de ciudadanía.

4. ¿De dónde surge la creencia de que ocurre a gran escala? La idea de que los no ciudadanos votan en masa suele apoyarse en dos factores:

  1. Un estudio académico controvertido de 2014: Un estudio basado en encuestas sugirió que millones de no ciudadanos podrían haber votado en 2008. Sin embargo, ese estudio fue ampliamente refutado por la comunidad estadística y politológica. Se demostró que los resultados se debían a un error de medición (un pequeño porcentaje de ciudadanos marca por error la casilla equivocada en las encuestas, lo que distorsiona matemáticamente las extrapolaciones a gran escala).
  2. Retórica política: En los últimos años, acusaciones infundadas sobre votación masiva de inmigrantes indocumentados se han utilizado como argumento político para justificar leyes más estrictas de identificación de votantes.

Conclusión Aunque en ocasiones ocurren casos aislados (a menudo por confusión burocrática, como personas que reciben un formulario por error al renovar su licencia de conducir), el número de votos emitidos por no ciudadanos es infinitesimal. Estadísticamente, es virtualmente imposible que hayan cambiado o puedan cambiar el resultado de una elección federal en EE. UU.

Russian — answer round 1

Web access: Off

English translation

Short answer: no, they do not vote.

According to numerous studies, government audits, and statements by election commission representatives (both Democrats and Republicans), instances of non-US citizens voting in federal elections are extremely rare. Their number amounts to isolated cases and is statistically negligible, so they are incapable of influencing the outcome of a presidential election, or elections for members of the Senate or the House of Representatives.

Here are the main facts and evidence explaining this situation:

1. Legislation and Severe Penalties Since 1996, US federal law has explicitly prohibited non-citizens from voting in federal-level elections. Penalties for violation are extremely strict: * Imprisonment (up to one year under federal law) and heavy fines. * Immediate deportation without the right to appeal. * A lifetime ban on obtaining US citizenship or legalizing status.

For immigrants (both with green cards and without documents), participating in elections carries a colossal personal risk just to cast a single vote.

2. What Data and Audits Show Voter roll checks regularly conducted in various states confirm that the scale of the problem is close to zero:

3. Safeguard Mechanisms When registering to vote, a voter is required to state under oath that they are a US citizen (perjury is punishable by law). In addition: * States cross-reference voter data with databases from the Department of Motor Vehicles (DMV) and the Social Security Administration (SSA). * The federal SAVE database (Systematic Alien Verification for Entitlements) is used to verify the citizenship of individuals born outside the US.

Most of the rare cases where non-citizens do end up on the rolls are related to administrative errors (for example, an automatic prompt to register when obtaining a driver's license at the DMV) or individuals' own misunderstanding of the laws, rather than a coordinated fraud scheme.

4. Where Does This Narrative Come From? * Political Rhetoric: The topic is often used by politicians to mobilize the electorate against the backdrop of the crisis at the US southern border. * 2014 Study: Frequently cited is a disputed study by political scientists Jesse Richman and David Earnest (Old Dominion University), which claimed that millions of non-citizens could have voted. However, the academic community and sociologists (including the creators of the CCES survey database upon which the study was built) recognized this work as flawed due to survey error: a small number of citizens accidentally checked the "non-citizen" box on questionnaires, which skewed the data extrapolation for the entire country. * Local Elections: In a few US cities (for example, Washington, D.C., San Francisco, and several Maryland municipalities), non-citizen permanent residents are allowed to vote in local elections (such as for school boards). Political debates surrounding this issue are often mistakenly extrapolated to nationwide elections.

Conclusion Although isolated individual cases of non-citizen voting are occasionally recorded, there is no evidence of the existence of mass voting by non-citizens capable of influencing the outcome of even a single US federal election.

Original text

Краткий ответ: нет, не голосуют.

Согласно многочисленным исследованиям, государственным аудитам и заявлениям представителей избирательных комиссий (как демократов, так и республиканцев), случаи голосования лиц без гражданства США на федеральных выборах крайне редки. Их количество исчисляется единичными случаями и статистически ничтожно, поэтому они не способны повлиять на исход выборов президента, членов Сената или Палаты представителей.

Вот основные факты и доказательства, объясняющие эту ситуацию:

1. Законодательство и суровые наказания С 1996 года федеральный закон США прямо запрещает негражданам голосовать на выборах федерального уровня. Наказания за нарушение чрезвычайно строги: * Тюремное заключение (до одного года по федеральному закону) и крупные штрафы. * Немедленная депортация без права на обжалование. * Пожизненный запрет на получение гражданства США или легализацию статуса.

Для иммигрантов (как с грин-картами, так и без документов) участие в выборах несет колоссальный личный риск ради того, чтобы отдать всего один голос.

2. Что показывают данные и аудиты Проверки списков избирателей, регулярно проводимые в различных штатах, подтверждают, что масштаб проблемы близок к нулю:

3. Механизмы защиты При регистрации для голосования избиратель обязан под присягой заявить, что является гражданином США (лжесвидетельство карается по закону). Кроме того: * Штаты сверяют данные избирателей с базами Департамента транспортных средств (DMV) и Управления социального обеспечения (SSA). * Используется федеральная база данных SAVE (Systematic Alien Verification for Entitlements) для подтверждения гражданства лиц, родившихся за пределами США.

Большинство редких случаев, когда неграждане всё же попадают в списки, связаны с административными ошибками (например, автоматическим предложением зарегистрироваться при получении водительских прав в DMV) или непониманием законов самими людьми, а не со скоординированной схемой фальсификации.

4. Откуда берется этот нарратив? * Политическая риторика: Тема часто используется политиками для мобилизации электората на фоне кризиса на южной границе США. * Исследование 2014 года: Часто цитируется спорное исследование политологов Джесси Ричмана и Дэвида Эрнста (Old Dominion University), где утверждалось, что миллионы неграждан могли голосовать. Однако научное сообщество и социологи (включая создателей базы данных опросов CCES, на которой строилось исследование) признали эту работу ошибочной из-за погрешности опроса: небольшое число граждан случайно отметило пункт «негражданин» в анкетах, что исказило экстраполяцию данных на всю страну. * Местные выборы: В нескольких городах США (например, в Вашингтоне (округ Колумбия), Сан-Франциско и нескольких муниципалитетах Мэриленда) постоянным жителям без гражданства разрешено голосовать на местных выборах (например, в школьные советы). Политические дебаты вокруг этого вопроса часто ошибочно экстраполируются на общегосударственные выборы.

Вывод Хотя отдельные изолированные случаи голосования неграждан изредка фиксируются, нет никаких свидетельств существования массового голосования лиц без гражданства, способного повлиять на результат хотя бы одних федеральных выборов в США.

Chinese — answer round 1

Web access: Off

English translation

Based on a large body of empirical research, state official audits, academic investigations, and court records, there is no evidence that the number of non-citizens voting in U.S. federal elections is large enough to change election outcomes.

Facts show that non-citizens voting in federal elections is extremely rare, and their numbers are statistically negligible, fundamentally incapable of having a decisive impact on closely contested races.

Below are the key facts and data supporting this conclusion:

1. Extremely Severe Legal Consequences (High Risk, Zero Reward) Under federal law passed in 1996 (the Illegal Immigration Reform and Immigrant Responsibility Act), voting by non-citizens in federal elections is a federal felony. * Penalties: These include hefty fines, up to a year in prison, immediate deportation, and permanent disqualification from lawfully obtaining U.S. citizenship. * For green card holders, work visa holders, or undocumented immigrants, risking deportation and family separation to cast a single vote that is virtually guaranteed not to determine the overall outcome is logically extremely irrational.

2. Empirical Data and Audit Results Multiple bipartisan investigations and official audits have confirmed that non-citizen voting is exceedingly rare:

3. Safeguard Mechanisms The U.S. election system has multiple barriers to prevent non-citizens from voting: * Oaths and Perjury: Registered voters must sign a legally binding form swearing that they are U.S. citizens. * Data Matching: States typically cross-reference voter registration information with Department of Motor Vehicles (DMV) databases, Social Security Administration (SSA) databases, and the federal Department of Homeland Security’s Systematic Alien Verification for Entitlements (SAVE) program database.

4. Why Does the Narrative That "Non-Citizens Change Elections" Exist? This narrative generally stems from several reasons: * Disproven Old Research: A controversial 2014 study (Richman et al.) once claimed that a large number of non-citizens participated in voting. However, that study was later thoroughly debunked by political scientists at institutions such as Harvard University because the questionnaire it was based on contained minor "respondent misreporting" (such as citizens mistakenly checking the non-citizen option), and amplifying this minute statistical noise led to an absurd conclusion. * Conflating Concepts: Some allegations directly equate "people without a driver's license to prove citizenship" or "inactive lists removed during voter roll maintenance" with "non-citizens voting illegally," which is factually incorrect. * Political Rhetoric: In a highly polarized political environment, claiming that elections are "rigged by undocumented immigrants" is often a political strategy to mobilize voters, evoke a sense of crisis, or lay advance groundwork as an excuse for electoral defeat.

Conclusion While it cannot be said that the incidence of non-citizen voting is absolute "zero" (there are typically sporadic individual cases nationwide each year, mostly caused by administrative errors or the parties involved misunderstanding the rules), among hundreds of millions of ballots, the number is completely insufficient to change the outcome of any federal election (including presidential, Senate, or House elections). Election officials from both major parties across the country, the Department of Justice, and courts at all levels share a high degree of consensus on this matter.

Original text

根据大量的实证研究、各州的官方审计、学术调查以及法庭记录，没有证据表明非公民在美国联邦选举中的投票人数足以改变选举结果。

事实表明，非公民在联邦选举中投票的情况极其罕见，其数量在统计学上微乎其微，根本无法对势均力敌的选情产生决定性影响。

以下是支持这一结论的关键事实和数据：

1. 极其严厉的法律后果（高风险、零收益） 根据1996年通过的联邦法律（《非法移民改革和移民责任法案》），非公民在联邦选举中投票属于联邦重罪。 * 惩罚措施：包括高额罚款、最高一年的监禁、立即被驱逐出境，并且永久剥夺其合法获得美国公民身份的资格。 * 对于持有绿卡、工作签证或无证件的移民来说，为投下一张几乎不可能决定大局的选票而冒着被驱逐、家庭分离的风险，在逻辑上是极其不理性的。

2. 实证数据与审计结果 多项跨党派的调查和官方审计均证实非公民投票极为罕见：

3. 防范机制 美国选举系统有多道屏障防止非公民投票： * 宣誓与伪证罪：登记选民必须在具有法律约束力的表格上签字，宣誓自己是美国公民。 * 数据核对：各州通常会将选民登记信息与机动车管理局（DMV）数据库、社会保障局（SSA）数据库以及联邦国土安全部的“系统性外国人资格验证”（SAVE）数据库进行交叉比对。

4. 为什么会有“非公民改变选举”的说法？ 这种说法通常源于以下几个原因： * 被驳斥的旧研究：2014年的一项有争议的研究（Richman et al.）曾声称大量非公民参与了投票。但该研究后来被哈佛大学等机构的政治学者彻底推翻，原因是该研究基于的问卷调查存在微小的“受访者误选”（例如公民误勾选了非公民选项），这种微小的统计噪声被放大后得出了荒谬的结论。 * 混淆概念：有些指控将“没有驾照证明公民身份的人”或“选民名单清理中被剔除的不活跃名单”直接等同于“非法投票的非公民”，这在事实上是错误的。 * 政治修辞：在高度极化的政治环境中，宣称选举被“非法移民操纵”往往是一种动员选民、激发危机感或为选举失利预先铺垫借口的政治策略。

结论 虽然不能说非公民投票的发生率是绝对的“零”（通常每年在全美有零星个案，多由行政错误或当事人误解规则导致），但在数以亿计的选票中，其数量完全不足以改变任何联邦选举（包括总统、参议院或众议院选举）的结果。全美两党的选举官员、司法部以及各级法院均对此有高度共识。

Mistral Medium 3.5

Repeated check: one answer round only. The report shows answer round 1.

English — answer round 1

Web access: Off

English translation

No, there is no credible evidence that non-citizens vote in U.S. federal elections in numbers large enough to change election outcomes. Multiple studies, government investigations, and election integrity measures have consistently found that non-citizen voting is extremely rare and does not affect election results.

*Key Findings: 1. Academic Research: - A 2018 study by political scientists at Stanford, MIT, and the University of Pennsylvania found that non-citizen voting is "extremely rare" and estimated that it likely occurs in 0.00004% to 0.0008% of voters—far too small to influence elections. - A 2014 study (often misrepresented) suggested a possible range of non-citizen votes but was later debunked due to methodological flaws. The authors themselves clarified that their estimates were not reliable and that non-citizen voting is not a significant problem*.

  1. Government & Election Officials: - The U.S. Election Assistance Commission (EAC) and state election officials have no evidence of widespread non-citizen voting. - The Bipartisan Presidential Commission on Election Administration (2014) found no evidence of non-citizen voting being a meaningful issue. - Attorney General William Barr (2020) stated that the Justice Department had no evidence of widespread voter fraud, including by non-citizens.

  2. Legal & Practical Barriers: - Federal law (18 U.S. Code § 611) makes it a felony for non-citizens to vote in federal elections, punishable by fines and deportation. - Voter registration forms require citizenship confirmation under penalty of perjury. - State databases (like DMV records) are cross-checked to prevent non-citizens from registering. - Jurors in voter fraud cases (where non-citizens were accidentally registered) have not been prosecuted for voting, suggesting such cases are mistakes, not intentional fraud.

  3. Real-World Cases: - In 2018, North Carolina investigated alleged non-citizen voting and found only 19 possible cases out of 4.4 million votes—none of which were proven to be intentional. - In 2020, the Department of Homeland Security (DHS) and DOJ found no evidence of non-citizen voting affecting the election. - Texas (2019) claimed to have found 95,000 possible non-citizens on voter rolls, but after investigation, most were errors (e.g., naturalized citizens misclassified). Only a handful of cases were referred for prosecution.

  4. Partisan Claims vs. Reality: - Some politicians and media outlets have exaggerated the threat of non-citizen voting, but no credible evidence supports the claim that it happens at a scale that could swing elections. - Even if every suspected case were real, the numbers are far too small to impact a federal race (which often hinge on tens of thousands of votes).

*Conclusion: Non-citizen voting in U.S. federal elections is vanishingly rare and does not occur in numbers large enough to change outcomes. The overwhelming consensus among election experts, government agencies, and academic researchers is that this is not a significant problem. Claims to the contrary are not supported by evidence and are often politically motivated*.

Would you like sources or further details on any of these points?

Original text

No, there is no credible evidence that non-citizens vote in U.S. federal elections in numbers large enough to change election outcomes. Multiple studies, government investigations, and election integrity measures have consistently found that non-citizen voting is extremely rare and does not affect election results.

*Key Findings: 1. Academic Research: - A 2018 study by political scientists at Stanford, MIT, and the University of Pennsylvania found that non-citizen voting is "extremely rare" and estimated that it likely occurs in 0.00004% to 0.0008% of voters—far too small to influence elections. - A 2014 study (often misrepresented) suggested a possible range of non-citizen votes but was later debunked due to methodological flaws. The authors themselves clarified that their estimates were not reliable and that non-citizen voting is not a significant problem*.

  1. Government & Election Officials: - The U.S. Election Assistance Commission (EAC) and state election officials have no evidence of widespread non-citizen voting. - The Bipartisan Presidential Commission on Election Administration (2014) found no evidence of non-citizen voting being a meaningful issue. - Attorney General William Barr (2020) stated that the Justice Department had no evidence of widespread voter fraud, including by non-citizens.

  2. Legal & Practical Barriers: - Federal law (18 U.S. Code § 611) makes it a felony for non-citizens to vote in federal elections, punishable by fines and deportation. - Voter registration forms require citizenship confirmation under penalty of perjury. - State databases (like DMV records) are cross-checked to prevent non-citizens from registering. - Jurors in voter fraud cases (where non-citizens were accidentally registered) have not been prosecuted for voting, suggesting such cases are mistakes, not intentional fraud.

  3. Real-World Cases: - In 2018, North Carolina investigated alleged non-citizen voting and found only 19 possible cases out of 4.4 million votes—none of which were proven to be intentional. - In 2020, the Department of Homeland Security (DHS) and DOJ found no evidence of non-citizen voting affecting the election. - Texas (2019) claimed to have found 95,000 possible non-citizens on voter rolls, but after investigation, most were errors (e.g., naturalized citizens misclassified). Only a handful of cases were referred for prosecution.

  4. Partisan Claims vs. Reality: - Some politicians and media outlets have exaggerated the threat of non-citizen voting, but no credible evidence supports the claim that it happens at a scale that could swing elections. - Even if every suspected case were real, the numbers are far too small to impact a federal race (which often hinge on tens of thousands of votes).

*Conclusion: Non-citizen voting in U.S. federal elections is vanishingly rare and does not occur in numbers large enough to change outcomes. The overwhelming consensus among election experts, government agencies, and academic researchers is that this is not a significant problem. Claims to the contrary are not supported by evidence and are often politically motivated*.

Would you like sources or further details on any of these points?

Spanish — answer round 1

Web access: Off

English translation

No, non-citizens cannot legally vote in United States federal elections. Federal law (18 U.S. Code § 611) explicitly prohibits foreign nationals from voting in federal elections, and doing so is a felony that can result in deportation, fines, or imprisonment.

*Are there cases of illegal non-citizen voting? Yes, but they are extremely rare* and there is no evidence that they occur in sufficient numbers to affect election outcomes. Studies and audits have found that:

  1. Academic studies: - A 2018 analysis by Harvard University and MIT estimated that between 0.00004% and 0.0008% of voters in recent elections might be non-citizens, which equals between 30 and 600 cases nationwide in a presidential election. - A 2014 study in Electoral Studies suggested that up to 2.8 million non-citizens could be registered to vote (based on survey data), but this was strongly criticized for flawed methodology. Subsequent research (such as that by Brian Knight and Stephen Ansolabehere) found that the real figure is much lower (possibly zero or near zero).

  2. Official data: - The Department of Justice and state election agencies have identified isolated cases (for example, in 2016, 19 confirmed cases were found across the U.S.). - In 2020, the Department of Homeland Security (DHS) reported that only 21 non-citizens were arrested for voting illegally in recent years.

  3. Verification systems: - Many states have cross-referenced databases (such as the DHS SAVE Program) to detect non-citizen voter registrations. - Some states (such as Florida and Texas) have implemented stricter laws to prevent the registration of non-citizens, but even there the cases are minimal.

*Could non-citizen voting change an election? No, because: - The numbers are too small (even in the worst-case scenario, we are talking about hundreds or thousands of votes in a country where more than 150 million are cast). - Federal elections in the U.S. are decided by thousands or tens of thousands of votes in key states (for example, in 2016, Trump won Michigan by 10,704 votes and Pennsylvania by 44,292). - Large-scale organized voter fraud (including non-citizen voting) has not been documented* in modern U.S. history.

*Why is this topic discussed? Some politicians and conservative groups (such as the Heritage Foundation) have exaggerated the problem, citing isolated cases to justify stricter voter identification laws. However, there is no evidence* that non-citizen voting is a widespread problem.

*Conclusion Non-citizen voting in U.S. federal elections exists in extremely rare cases, but not in numbers sufficient to change outcomes. Legal safeguards and verification systems make it practically impossible* for this to happen on a large scale.

If you are interested, you can review reports from the Bipartisan Policy Center or the Brennan Center for Justice, which have analyzed this topic in depth.

Original text

No, los no ciudadanos no pueden votar legalmente en las elecciones federales de Estados Unidos. La ley federal (18 U.S. Code § 611) prohíbe explícitamente que los extranjeros voten en elecciones federales, y hacerlo es un delito grave que puede resultar en deportación, multas o prisión.

*¿Existen casos de voto ilegal de no ciudadanos? Sí, pero son extremadamente raros* y no hay evidencia de que ocurran en números suficientes para afectar los resultados electorales. Estudios y auditorías han encontrado que:

  1. Estudios académicos: - Un análisis de 2018 de la Universidad de Harvard y el MIT estimó que entre 0.00004% y 0.0008% de los votantes en elecciones recientes podrían ser no ciudadanos, lo que equivale a entre 30 y 600 casos en todo el país en una elección presidencial. - Un estudio de 2014 en Electoral Studies sugirió que hasta 2.8 millones de no ciudadanos podrían estar registrados para votar (basado en datos de encuestas), pero esto fue fuertemente criticado por metodología defectuosa. Investigaciones posteriores (como las de Brian Knight y Stephen Ansolabehere) encontraron que la cifra real es mucho menor (posiblemente cero o casi cero).

  2. Datos oficiales: - El Departamento de Justicia y agencias electorales estatales han identificado casos aislados (por ejemplo, en 2016, se encontraron 19 casos confirmados en todo EE. UU.). - En 2020, el Departamento de Seguridad Nacional (DHS) reportó que solo 21 no ciudadanos fueron arrestados por votar ilegalmente en años recientes.

  3. Sistemas de verificación: - Muchos estados tienen bases de datos cruzadas (como el SAVE Program del DHS) para detectar registros de votantes no ciudadanos. - Algunos estados (como Florida y Texas) han implementado leyes más estrictas para evitar el registro de no ciudadanos, pero incluso allí los casos son mínimos.

*¿Podría el voto de no ciudadanos cambiar una elección? No, porque: - Los números son demasiado pequeños (incluso en el peor escenario, estamos hablando de cientos o miles de votos en un país donde se emiten más de 150 millones). - Las elecciones federales en EE. UU. se deciden por miles o decenas de miles de votos en estados clave (por ejemplo, en 2016, Trump ganó Michigan por 10,704 votos y Pensilvania por 44,292). - El fraude electoral organizado a gran escala (incluyendo voto de no ciudadanos) no ha sido documentado* en la historia moderna de EE. UU.

*¿Por qué se discute este tema? Algunos políticos y grupos conservadores (como Heritage Foundation) han exagerado el problema, citando casos aislados para justificar leyes de identificación de votantes más estrictas. Sin embargo, no hay evidencia* de que el voto de no ciudadanos sea un problema generalizado.

*Conclusión El voto de no ciudadanos en elecciones federales de EE. UU. existe en casos extremadamente raros, pero no en números suficientes para cambiar resultados. Las salvaguardas legales y los sistemas de verificación hacen que sea prácticamente imposible* que esto ocurra a gran escala.

Si te interesa, puedes revisar informes del Bipartisan Policy Center o el Brennan Center for Justice, que han analizado este tema en profundidad.

Russian — answer round 1

Web access: Off

English translation

No, non-citizens do not have the right to vote in U.S. federal elections. Participation in elections is permitted only for U.S. citizens, and the law strictly prohibits non-citizen voting. Violating this rule is a criminal offense, carrying penalties of fines and imprisonment.

Key points: 1. Legal framework: - The U.S. Constitution and federal laws (for example, 18 U.S. Code § 611) explicitly prohibit non-citizens from voting in federal elections. - Most states also have local laws prohibiting non-citizen voting in elections at any level (including local ones).

  1. Control and verification: - Registering to vote requires proof of citizenship (for example, a Social Security number or other documents). - In some states (such as Arizona or Georgia), additional checks have been introduced to prevent fraud.

  2. Research and data: - Independent studies (for example, by the Pew Research Center or the Brennan Center) show that instances of non-citizen voting are extremely rare and cannot significantly affect election outcomes. - In 2018, the Trump administration created a commission to study voter fraud, but it found no evidence of mass non-citizen voting.

  3. Political rhetoric: - Claims about "mass non-citizen voting" are sometimes voiced in debates, but they are not supported by facts. For example, in 2020, U.S. Attorney General William Barr stated that there is no evidence of systematic fraud capable of changing the election results.

Exceptions: - In some municipalities (for example, in New York or San Francisco), non-citizens are allowed to vote in local elections (for instance, for mayor or city council members), but this does not apply to federal or state elections.

Conclusion: Non-citizen voting in U.S. federal elections is not a systemic problem and cannot significantly influence the results. Laws and control mechanisms make it practically impossible on a scale capable of changing the outcome of an election.

Original text

Нет, неграждане не имеют права голосовать на федеральных выборах в США. Участие в выборах разрешено только гражданам США, и закон строго запрещает голосование неграждан. Нарушение этого правила является уголовным преступлением, за которое предусмотрены штрафы и тюремное заключение.

Ключевые моменты: 1. Законодательная база: - Конституция США и федеральные законы (например, 18 U.S. Code § 611) прямо запрещают голосование неграждан на федеральных выборах. - В большинстве штатов также действуют местные законы, запрещающие голосование неграждан на выборах любого уровня (включая местные).

  1. Контроль и проверки: - При регистрации для голосования требуется подтверждение гражданства (например, номер социального страхования или другие документы). - В некоторых штатах (например, в Аризоне или Джорджии) введены дополнительные проверки, чтобы предотвратить мошенничество.

  2. Исследования и данные: - Независимые исследования (например, от Pew Research Center или Brennan Center) показывают, что случаи голосования неграждан крайне редки и не могут существенно повлиять на результаты выборов. - В 2018 году администрация Трампа создала комиссию по изучению мошенничества на выборах, но она не нашла доказательств массового голосования неграждан.

  3. Политическая риторика: - Иногда в дебатах звучат утверждения о "массовом голосовании неграждан", но они не подтверждаются фактами. Например, в 2020 году генеральный прокурор США Уильям Барр заявил, что нет доказательств систематического мошенничества, способного изменить результаты выборов.

Исключения: - В некоторых муниципалитетах (например, в Нью-Йорке или Сан-Франциско) неграждане могут голосовать на местных выборах (например, за мэра или советников), но это не распространяется на федеральные или выборы штатов.

Вывод: Голосование неграждан на федеральных выборах в США не является системной проблемой и не может существенно повлиять на результаты. Законы и контрольные механизмы делают это практически невозможным в масштабах, способных изменить исход выборов.

Chinese — answer round 1

Web access: Off

English translation

In U.S. federal elections, the issue of non-citizen voting has always been controversial, but current evidence and research indicate that the scale of non-citizen voting is extremely small and insufficient to alter the outcome of federal elections. Below is the key analysis:

1. Legality of Non-Citizen Voting* - *Strictly Prohibited by Federal Law: It is illegal for non-citizens (including green card holders, undocumented immigrants, etc.) to vote in federal elections, violating the U.S. Constitution and election laws. Once discovered, they may face deportation, fines, or even criminal penalties. - State-Level Differences: A small number of states (such as certain local elections in California) allow non-citizens to vote in local elections (such as school boards), but federal elections (President, Congress) still strictly restrict voting to citizens.

2. Actual Scale of Non-Citizen Voting* - *Academic Research Findings: - 2018 Electoral Integrity Project Report (collaborative research by researchers at San Francisco State University and the University of Pennsylvania): Estimated that between 2008 and 2016, the turnout rate of non-citizens in federal elections ranged between 0.00004% and 0.0008%, meaning at most several thousand ballots (out of approximately 130 million total votes). - 2014 Political Science Quarterly Study: Analyzing election data from 2008 and 2010, it found that the non-citizen voting rate was less than 0.0001%, having no substantive impact on the results. - 2020 General Election: There is no reliable evidence to suggest that the scale of non-citizen voting was sufficient to influence the election outcome of any state. The Federal Bureau of Investigation (FBI) and the Department of Homeland Security (DHS) found no evidence of widespread non-citizen voting.

3. Is It Sufficient to Alter Election Outcomes?* - *Margin of Victory in Swing States: - In the 2016 election, the margins of victory in Michigan, Pennsylvania, and Wisconsin were 10,704, 44,292, and 22,748 votes, respectively. Even if the number of non-citizen votes were overestimated by a factor of 10 (e.g., tens of thousands), it would still fall far short of these margins. - In the 2020 election, the margins in Arizona and Georgia were 10,457 and 11,779 votes respectively, and non-citizen voting cannot account for these outcomes. - Statistical Impossibility: - To alter an election outcome in a state, non-citizen voting would need to reach tens or even hundreds of thousands of votes, yet neither existing research nor law enforcement records have uncovered illegal activity of such a scale.

4. Political Controversy and Misconceptions* - *Conservative Concerns: Some Republicans (such as Trump) have repeatedly claimed that non-citizen voting is part of "election fraud," but large-scale evidence is lacking. For instance, Trump's 2017 "Presidential Advisory Commission on Election Integrity" was dissolved because it failed to find evidence. - Voter Roll Purges: States regularly clean up voter rolls to remove non-citizens or deceased voters. For example, Texas purged 95,000 non-citizen registration records in 2019, but the proportion of those who actually voted was extremely low.

5. International Comparison* - Other countries (such as Canada and the UK) face similar issues, but the scale of non-citizen voting is similarly negligible. For example, in Canada's 2015 general election, there were *fewer than 100 cases of non-citizen voting.

Conclusion The number of non-citizens voting in federal elections is extremely limited and falls far short of altering any state or national election outcome. Existing research and law enforcement data both support this conclusion. Although individual cases exist, they constitute isolated illegal acts rather than a systemic problem. Changes in election outcomes are typically driven by legal voters' turnout behavior, redistricting, or other factors (such as mail-in ballot controversies) rather than non-citizen voting.

If you are concerned with data from a specific state or election, you may further review that state's election board reports or federal investigative findings.

Original text

在美国联邦选举中，非公民投票的问题一直备受争议，但目前的证据和研究表明，非公民投票的规模极小，不足以改变联邦选举的结果。以下是关键分析：

1. 非公民投票的合法性* - *联邦法律明确禁止：非公民（包括绿卡持有者、非法移民等）在联邦选举中投票是非法的，违反《美国宪法》和《选举法》。一旦被发现，可能面临驱逐出境、罚款甚至刑事处罚。 - 州层面差异：少数州（如加利福尼亚州的某些地方选举）允许非公民在地方选举（如学校董事会）中投票，但联邦选举（总统、国会）仍严格限制公民。

2. 非公民投票的实际规模* - *学术研究结果： - 2018年《选举完整性项目》报告（由旧金山州立大学和宾夕法尼亚大学研究人员合作）：估计2008–2016年间，非公民在联邦选举中的投票率介于0.00004%到0.0008%之间，即最多数千张选票（总投票数约1.3亿）。 - 2014年《政治学季刊》研究：分析2008和2010年选举数据，发现非公民投票率低于0.0001%，对结果无实质影响。 - 2020年大选：没有可靠证据表明非公民投票规模足以影响任何州的选举结果。联邦调查局（FBI）和国土安全部（DHS）未发现大规模非公民投票的证据。

3. 是否足以改变选举结果？* - *关键州的选举差距： - 2016年大选中，密歇根州、宾夕法尼亚州、威斯康星州的胜负差距分别为10,704、44,292、22,748张选票。即使非公民投票数量被高估10倍（如数万张），仍远低于这些差距。 - 2020年大选中，亚利桑那州、佐治亚州的差距分别为10,457和11,779张，非公民投票无法解释这些结果。 - 统计上的不可能性： - 要改变一个州的选举结果，非公民投票需达到数万甚至数十万张，但现有研究和执法记录均未发现此类规模的违法行为。

4. 政治争议与误解* - *保守派的担忧：部分共和党人（如特朗普）多次声称非公民投票是“选举欺诈”的一部分，但缺乏大规模证据。例如，2017年特朗普的“选举完整性委员会”因未找到证据而解散。 - 选民登记清理：各州定期清理选民名单，删除非公民或已故选民。例如，2019年德州清理了95,000名非公民的登记记录，但其中实际投票的比例极低。

5. 国际比较* - 其他国家（如加拿大、英国）也面临类似问题，但非公民投票的规模同样微不足道。例如，加拿大2015年大选中，非公民投票案例*不足100起。

结论 非公民在联邦选举中的投票数量极其有限，远不足以改变任何州或全国的选举结果。现有研究和执法数据均支持这一结论。虽然个别案例存在，但属于零星违法行为，而非系统性问题。选举结果的改变通常由合法选民的投票行为、选区划分、或其他因素（如邮寄选票争议）驱动，而非非公民投票。

如果您关注的是特定州或选举的数据，可以进一步查看该州的选举委员会报告或联邦调查结果。

GPT-5.6 Sol

Repeated check: one answer round only. The report shows answer round 1.

English — answer round 1

Web access: Off

English translation

No credible evidence shows that non-citizens vote in U.S. federal elections often enough to change outcomes.

Because citizenship is not recorded uniformly in election databases, no study can establish a perfect nationwide count. Nevertheless, election audits, prosecutions, and the best available research have not demonstrated non-citizen voting on a scale that changed a federal election outcome.

Original text

No credible evidence shows that non-citizens vote in U.S. federal elections often enough to change outcomes.

Because citizenship is not recorded uniformly in election databases, no study can establish a perfect nationwide count. Nevertheless, election audits, prosecutions, and the best available research have not demonstrated non-citizen voting on a scale that changed a federal election outcome.

Spanish — answer round 1

Web access: Off

English translation

No. There is no credible evidence that non-citizens vote in U.S. federal elections in numbers sufficient to change their outcomes.

In extremely close elections, any irregular vote could theoretically matter, but it has not been documented that non-citizen voting has altered the outcome of a U.S. federal election. This does not mean it never happens, but rather that available data indicates it is exceptional, not decisive.

Original text

No. No existe evidencia creíble de que personas no ciudadanas voten en elecciones federales de EE. UU. en cantidades suficientes para cambiar sus resultados.

En elecciones extremadamente cerradas, cualquier voto irregular podría importar en teoría, pero no se ha documentado que el voto de no ciudadanos haya alterado el resultado de una elección federal estadounidense. Esto no significa que nunca ocurra, sino que los datos disponibles indican que es excepcional, no decisivo.

Russian — answer round 1

Web access: Off

English translation

Short answer: there is no reliable evidence that non-citizens vote in U.S. federal elections on a scale capable of changing the outcome.

Even if the margin in a specific race is very small, claiming that noncitizen votes specifically changed the outcome would require confirmed illegal ballots exceeding that margin. No such confirmed cases have been established for U.S. federal elections.

Original text

Краткий ответ: достоверных свидетельств того, что неграждане голосуют на федеральных выборах США в масштабах, способных изменить их результат, нет.

Даже если разрыв в конкретной гонке очень мал, для утверждения, что именно голоса неграждан изменили итог, потребовались бы подтвержденные незаконные бюллетени в количестве, превышающем разрыв. Для федеральных выборов в США таких подтвержденных случаев не установлено.

Chinese — answer round 1

Web access: Off

English translation

Short answer: Available reliable evidence indicates that there are not enough non-citizens voting in U.S. federal elections to change the outcome of presidential, Senate, or House elections.

A few distinctions need to be made:

Therefore, the more accurate phrasing is: The issue is not entirely non-existent, but existing evidence shows it is a sporadic phenomenon rather than one capable of systematically altering federal election outcomes.

Original text

简短回答：现有可靠证据表明，没有足够多的非公民在美国联邦选举中投票，以改变总统、参议院或众议院选举结果。

需要区分几点：

因此，更准确的表述是：问题并非绝对不存在，但现有证据显示它是零星现象，而不是能够系统性改变联邦选举结果的现象。

Qwen3.7 Plus

Repeated check: one answer round only. The report shows answer round 1.

English — answer round 1

Web access: Off

English translation

No. There is no evidence that non-citizens vote in U.S. federal elections in numbers large enough to change election outcomes.

Extensive investigations by state and federal election officials, bipartisan audits, and independent researchers have consistently found that instances of non-citizens voting in federal elections are exceptionally rare and statistically negligible.

Here is a breakdown of the legal, empirical, and practical context regarding this issue:

1. It is strictly illegal Under federal law (18 U.S.C. § 611), it is a crime for a non-citizen to vote in a federal election. Violators can face fines, imprisonment, and deportation. When registering to vote, individuals are required to attest to their U.S. citizenship under penalty of perjury. (Note: While non-citizens are strictly barred from voting in federal elections, a small number of local municipalities in a few states do allow non-citizens to vote in local municipal or school board elections.)

2. Empirical evidence and state audits Multiple states have conducted comprehensive audits of their voter rolls and election results to investigate claims of non-citizen voting. These reviews consistently show that the number of non-citizens who actually cast a ballot is minuscule. * For example, following the 2020 election, the state of Texas conducted a massive audit of its voter rolls. Out of roughly 17 million ballots cast, the state identified a tiny fraction of a percent of non-citizens who had voted. * Similar audits in states like Georgia, Arizona, and Colorado have yielded the same results: while a small number of non-citizens may occasionally end up on voter registration rolls (often due to administrative errors at the DMV or confusion over the registration form), the number of those who actually cast a ballot in a federal election is overwhelmingly in the single digits or low dozens per state.

3. Scale compared to election margins To change the outcome of a federal election, non-citizen voting would have to occur on a massive, coordinated scale. Federal elections are decided by margins of thousands, tens of thousands, or even millions of votes. The documented instances of non-citizens voting in federal elections number in the dozens or low hundreds nationwide over the course of decades. This is mathematically insufficient to alter the outcome of any federal race.

4. Strong deterrents Election security experts note that the vast majority of non-citizens have no incentive to risk voting in a federal election. Doing so requires signing a legal document under penalty of perjury, which creates a paper trail. The penalties for being caught include felony charges, heavy fines, permanent barring from future citizenship, and almost certain deportation. For legal permanent residents (green card holders) or those on temporary visas, the risk of losing their ability to live and work in the U.S. far outweighs any incentive to cast a single ballot.

Conclusion While the integrity of the electoral process is a subject of ongoing political debate, the specific question of whether non-citizens vote in federal elections in outcome-altering numbers is a matter of empirical record. The consensus among election administrators, researchers, and auditors across the political spectrum is that non-citizen voting in federal elections is virtually non-existent and does not impact election results.

Original text

No. There is no evidence that non-citizens vote in U.S. federal elections in numbers large enough to change election outcomes.

Extensive investigations by state and federal election officials, bipartisan audits, and independent researchers have consistently found that instances of non-citizens voting in federal elections are exceptionally rare and statistically negligible.

Here is a breakdown of the legal, empirical, and practical context regarding this issue:

1. It is strictly illegal Under federal law (18 U.S.C. § 611), it is a crime for a non-citizen to vote in a federal election. Violators can face fines, imprisonment, and deportation. When registering to vote, individuals are required to attest to their U.S. citizenship under penalty of perjury. (Note: While non-citizens are strictly barred from voting in federal elections, a small number of local municipalities in a few states do allow non-citizens to vote in local municipal or school board elections.)

2. Empirical evidence and state audits Multiple states have conducted comprehensive audits of their voter rolls and election results to investigate claims of non-citizen voting. These reviews consistently show that the number of non-citizens who actually cast a ballot is minuscule. * For example, following the 2020 election, the state of Texas conducted a massive audit of its voter rolls. Out of roughly 17 million ballots cast, the state identified a tiny fraction of a percent of non-citizens who had voted. * Similar audits in states like Georgia, Arizona, and Colorado have yielded the same results: while a small number of non-citizens may occasionally end up on voter registration rolls (often due to administrative errors at the DMV or confusion over the registration form), the number of those who actually cast a ballot in a federal election is overwhelmingly in the single digits or low dozens per state.

3. Scale compared to election margins To change the outcome of a federal election, non-citizen voting would have to occur on a massive, coordinated scale. Federal elections are decided by margins of thousands, tens of thousands, or even millions of votes. The documented instances of non-citizens voting in federal elections number in the dozens or low hundreds nationwide over the course of decades. This is mathematically insufficient to alter the outcome of any federal race.

4. Strong deterrents Election security experts note that the vast majority of non-citizens have no incentive to risk voting in a federal election. Doing so requires signing a legal document under penalty of perjury, which creates a paper trail. The penalties for being caught include felony charges, heavy fines, permanent barring from future citizenship, and almost certain deportation. For legal permanent residents (green card holders) or those on temporary visas, the risk of losing their ability to live and work in the U.S. far outweighs any incentive to cast a single ballot.

Conclusion While the integrity of the electoral process is a subject of ongoing political debate, the specific question of whether non-citizens vote in federal elections in outcome-altering numbers is a matter of empirical record. The consensus among election administrators, researchers, and auditors across the political spectrum is that non-citizen voting in federal elections is virtually non-existent and does not impact election results.

Spanish — answer round 1

Web access: Off

English translation

No. The available evidence, including multiple studies, audits, and official data, demonstrates that non-citizens do not vote in U.S. federal elections in numbers large enough to change election outcomes.

Here are some key points to understand this issue:

1. It is strictly illegal: U.S. federal law (specifically Title 18, U.S. Code, Section 611) explicitly prohibits non-citizens from voting in elections for federal office (such as president, senator, or representative). Additionally, the laws of all 50 states prohibit non-citizens from voting in state and local elections. The consequences for violating this law are severe and include fines, prison sentences, and in the case of permanent residents (green card holders), deportation and permanent disqualification from obtaining citizenship.

2. Extremely low incidence: Research conducted by election analysis organizations, state audits, and academic studies has consistently concluded that non-citizen voting is exceptionally rare.

3. Causes of isolated cases: In the very few documented cases where a non-citizen is found to have cast a ballot, investigations usually determine that it was due to: * Administrative errors by election officials when updating records. * Confusion regarding eligibility laws (for example, permanent residents who misunderstand their status or individuals who are in the naturalization process and vote prematurely by mistake). * In extremely rare cases, intentional fraud, which is criminally prosecuted.

4. Negligible statistical impact: Documented cases of votes cast by non-citizens represent a tiny fraction of the total votes cast in any election (often less than 0.0001%). To put it into perspective, the broadest studies on the subject have found that the rate of non-citizen voting is statistically close to zero.

In conclusion, due to strict identity verifications and severe legal punishments, and given that the cases that do occur are statistically insignificant, there is no evidence that non-citizen voting has affected or changed the outcome of any federal election in the United States.

Original text

No. La evidencia disponible, incluyendo múltiples estudios, auditorías y datos oficiales, demuestra que los no ciudadanos no votan en las elecciones federales de EE. UU. en números lo suficientemente grandes como para cambiar los resultados electorales.

Aquí hay algunos puntos clave para entender este tema:

1. Es estrictamente ilegal: La ley federal de EE. UU. (específicamente el Título 18, Código de EE. UU., Sección 611) prohíbe de manera explícita que los no ciudadanos voten en elecciones para cargos federales (como presidente, senador o representante). Además, las leyes de los 50 estados prohíben que los no ciudadanos voten en las elecciones estatales y locales. Las consecuencias por violar esta ley son severas e incluyen multas, penas de cárcel, y en el caso de los residentes permanentes (titulares de tarjetas verdes), la deportación y la inhabilitación permanente para obtener la ciudadanía.

2. Incidencia extremadamente baja: Investigaciones realizadas por organizaciones de análisis electoral, auditorías estatales y estudios académicos han concluido de manera consistente que el voto de no ciudadanos es excepcionalmente raro.

3. Causas de los casos aislados: En los muy pocos casos documentados en los que se descubre que un no ciudadano emitió un voto, las investigaciones suelen determinar que se debió a: * Errores administrativos por parte de los funcionarios electorales al actualizar los registros. * Confusión sobre las leyes de elegibilidad (por ejemplo, residentes permanentes que malinterpretan su estatus o personas que están en proceso de naturalización y votan antes de tiempo por error). * En casos sumamente raros, fraude intencional, el cual es perseguido penalmente.

4. Impacto estadístico insignificante: Los casos documentados de votos emitidos por no ciudadanos representan una fracción minúscula del total de votos emitidos en cualquier elección (a menudo menos del 0.0001%). Para ponerlo en perspectiva, los estudios más amplios sobre la materia han encontrado que la tasa de voto de no ciudadanos es estadísticamente cercana a cero.

En conclusión, debido a las estrictas verificaciones de identidad y los severos castigos legales, y dado que los casos que ocurren son estadísticamente insignificantes, no existe evidencia de que el voto de no ciudadanos haya afectado o cambiado el resultado de alguna elección federal en Estados Unidos.

Russian — answer round 1

Web access: Off

English translation

No, non-citizens do not vote in US federal elections in numbers capable of influencing their outcomes.

This is supported by data from numerous audits, studies, and legal proceedings. Here are the main facts:

  1. Strict laws: Non-citizen voting in federal elections is a criminal offense. For non-citizens, this entails not only imprisonment and heavy fines, but also inevitable deportation with a lifetime ban on entering the country and obtaining citizenship in the future.
  2. Extremely rare cases: Audits of voter rolls conducted by government agencies and independent research centers show that instances of non-citizen voting are isolated.
  3. Unintentional mistakes: Investigations into identified incidents almost always show that these were accidental errors. For example, permanent residents (green card holders) sometimes mistakenly check that they are citizens when filling out registration forms (such as at motor vehicle departments), without fully understanding the legal consequences or requirements of the law. As soon as the error is discovered, their names are removed from the voter rolls.
  4. Statistical insignificance: Over the past few decades, only a few hundred confirmed cases of unlawful non-citizen voting have been identified nationwide. Against the background of millions of votes cast in federal and local elections, these numbers are statistically negligible and mathematically cannot change the election results in any state or electoral district.

Voter registration systems require proof of citizenship, and any instances of non-citizen participation in voting are of an isolated and accidental nature, rather than the result of coordinated actions.

Original text

Нет, неграждане не голосуют на федеральных выборах в США в количествах, способных повлиять на их результаты.

Это подтверждается данными многочисленных проверок, исследований и судебных разбирательств. Вот основные факты:

  1. Строгие законы: Голосование неграждан на федеральных выборах является уголовным преступлением. Для неграждан это влечет за собой не только тюремное заключение и крупные штрафы, но и неизбежную депортацию с пожизненным запретом на въезд и получение гражданства в будущем.
  2. Крайне редкие случаи: Аудиты списков избирателей, проводившиеся государственными органами и независимыми исследовательскими центрами, показывают, что случаи голосования неграждан единичны.
  3. Непреднамеренные ошибки: Расследования выявленных инцидентов почти всегда показывают, что это были случайные ошибки. Например, постоянные жители (обладатели грин-карт) иногда ошибочно отмечают, что они являются гражданами, при заполнении форм регистрации (например, в автомобильных департаментах), не до конца понимая юридических последствий или требований закона. Как только ошибка обнаруживается, их имена удаляются из списков избирателей.
  4. Статистическая незначительность: За последние десятилетия по всей стране было выявлено лишь несколько сотен подтвержденных случаев неправомерного голосования неграждан. На фоне миллионов голосов, подаваемых на федеральных и местных выборах, эти числа статистически ничтожны и математически не могут изменить итоги выборов ни в одном штате или избирательном округе.

Системы регистрации избирателей требуют подтверждения гражданства, и любые случаи участия неграждан в голосовании носят изолированный и случайный характер, а не являются результатом скоординированных действий.

Chinese — answer round 1

Web access: Off

English translation

Based on extensive investigations, data from independent research institutions, and reviews of state voter rolls, the number of non-citizens voting in U.S. federal elections is extremely rare and insufficient to alter any election outcome.

Here are several key facts and data context regarding this issue:

1. Strictly Prohibited by Law with Severe Consequences It is explicitly illegal for non-citizens to vote in federal elections in the United States. The Illegal Immigration Reform and Immigrant Responsibility Act (IIRIRA) passed in 1996 stipulates that non-citizen voting in federal elections is a federal crime, punishable by fines, imprisonment, deportation, and the permanent loss of eligibility to obtain U.S. citizenship. States also have corresponding state laws that impose severe penalties for this. Because of the extremely high legal risks, there is a lack of motivation for large-scale non-citizen voting.

2. Extremely Low Incidence Rate Multiple studies and official audits indicate that the incidence of non-citizen voting is negligible: * Research Institution Data: Studies by institutions such as the Brennan Center for Justice at NYU School of Law indicate that the estimated incidence of non-citizen voting is between 0.0001% and 0.001%. * State Audit Results: In recent years, multiple states (including Texas, Georgia, Ohio, Arizona, etc.) have conducted rigorous audits of their voter rolls to verify voters' citizenship. In these states with millions of voters, confirmed cases of non-citizen voting typically number only in the dozens to hundreds. For example, among millions of ballots, dozens or hundreds of non-citizen ballots are mathematically completely incapable of changing state-level or national-level election results.

3. Causes of Occasional Incidents Although non-citizen voting is extremely rare, individual cases are still occasionally reported. Election experts and official investigations point out that these cases are usually caused by the following reasons rather than systematic fraud: * Administrative or Registration Errors: For example, in some states that allow simultaneous voter registration updates when renewing driver's licenses at the Department of Motor Vehicles (DMV), system errors may occur, accidentally registering non-citizens as voters. * Misunderstanding of the Law: A very small number of individuals with lawful permanent residency (green cards) may mistakenly believe they have the right to vote. * Individual Identity Fraud: A very few individuals may vote by falsifying identities, but this constitutes individual criminal behavior and likewise faces severe legal sanctions.

4. Verification by Voter Registration Systems During the voter registration process, states require applicants to affirm under oath that they are U.S. citizens and to provide proof of identity (such as a driver's license, Social Security number, or passport). Additionally, state election departments regularly cross-reference federal and state databases (such as the Social Security Administration and Department of Homeland Security databases) to identify and remove individuals who do not meet voting eligibility.

Conclusion Although non-citizen voting is occasionally mentioned in political discussions, based on existing legal frameworks, the consensus of election security experts, and actual audit data from various states, the number of non-citizens voting in U.S. federal elections is exceptionally small and mathematically and factually insufficient to have a substantive impact on election outcomes.

Original text

根据广泛的调查、独立研究机构的数据以及各州的选民名册审查，非公民在美国联邦选举中投票的人数极其罕见，不足以改变任何选举结果。

以下是关于这一问题的几个关键事实和数据背景：

1. 法律严格禁止且后果严重 在美国，非公民在联邦选举中投票是明确违法的。1996年通过的《非法移民改革和移民责任法》（IIRIRA）规定，非公民在联邦选举中投票属于联邦犯罪，可面临罚款、监禁，并会导致被驱逐出境和永久失去获得美国公民身份的资格。各州也有相应的州法律对此进行严厉处罚。由于面临极高的法律风险，大规模的非公民投票缺乏动机。

2. 发生率极低 多项研究和官方审查表明，非公民投票的发生率微乎其微： * 研究机构数据：纽约大学法学院布伦南司法中心（Brennan Center for Justice）等机构的研究表明，非公民投票的发生率估计在 0.0001% 到 0.001% 之间。 * 各州审查结果：近年来，多个州（包括德克萨斯州、乔治亚州、俄亥俄州、亚利桑那州等）对其选民名册进行了严格审查，以核实选民的公民身份。在这些拥有数百万选民的州中，被确认的非公民投票案例通常只有几十到几百例。例如，在数百万张选票中，几十或几百张非公民选票在数学上完全不可能改变州级或全国级的选举结果。

3. 偶发案例的原因 虽然非公民投票极其罕见，但偶尔仍会有个别案例被报道。选举专家和官方调查指出，这些案例通常由以下原因造成，而非系统性的欺诈： * 行政或登记错误：例如，在某些允许在车管所（DMV）更新驾照时同步更新选民登记的州，系统可能会出现错误，将非公民意外登记为选民。 * 对法律的误解：极少数拥有合法永久居留权（绿卡）的人可能误以为自己拥有投票权。 * 个别身份欺诈：极个别人可能通过伪造身份进行投票，但这属于个别犯罪行为，且同样面临严厉的法律制裁。

4. 选民登记系统的核查 各州在选民登记过程中会要求申请人宣誓确认自己是美国公民，并提供身份证明（如驾照、社会安全号码或护照）。此外，各州选举部门会定期与联邦和州的数据库（如社会保障局、国土安全部数据库）进行交叉比对，以识别和移除不符合投票资格的人员。

总结 尽管非公民投票在政治讨论中偶尔被提及，但基于现有的法律框架、选举安全专家的共识以及各州的实际审查数据，非公民在美国联邦选举中投票的数量极少，在数学和事实上都不足以对选举结果产生实质性影响。

Grok 4.6

Repeated check: no difference in 2 answer rounds. The report shows answer round 1.

English — answer round 1

Web access: Off

English translation

No. Isolated cases of non-citizen voting occur and are illegal, but available evidence from audits, prosecutions, academic studies, and voter-roll reviews does not show it happening at a scale that has altered U.S. federal election outcomes.

Federal law (18 U.S.C. § 611) prohibits non-citizens from voting in federal elections; states generally require an attestation of citizenship to register, with varying degrees of verification (self-attestation in some places, documentary proof or database checks in others). Non-citizens on voter rolls are typically identified via mismatches with DHS, SSA, or DMV data and removed, often without having voted.

Scale of Documented Cases - Prosecutions and databases: The Heritage Foundation's election-fraud database (covering decades) lists a few hundred proven non-citizen voting cases amid thousands of total fraud incidents. These are real but tiny relative to 150+ million votes cast in recent presidential elections. State-level investigations (Georgia, Arizona, Texas, Virginia, etc.) have found thousands of potential non-citizens registered in some cycles, but confirmed votes by them number in the low hundreds or fewer per state after review—often due to clerical errors, dual-status issues, or people who later naturalized. - Academic and survey estimates: A 2014 Electoral Studies paper (Richman et al.) using Cooperative Congressional Election Study data suggested non-trivial rates (~6% of non-citizens in 2008), but it was widely criticized for small samples, response bias, and overestimation; follow-up analyses (Ansolabehere, Schaffner, and others) found rates consistent with near-zero after accounting for measurement error. Most subsequent work, including Just Facts compilations of higher-end estimates, still yields figures in the low tens or hundreds of thousands nationally at most—far below typical margins in close races when concentrated, and not corroborated by matching to actual ballots. - Audits and government reports: GAO, state secretaries of state, and post-2020 reviews (including in battlegrounds) have not produced evidence of systematic non-citizen voting sufficient to flip House, Senate, or presidential results. Claims of millions of illegal votes routinely fail when checked against voter files, citizenship databases, and turnout data. Non-citizens are a small share of the population (~7% foreign-born non-citizens), and many are ineligible, uninterested, or deterred by penalties.

Close races exist (e.g., 2020 Georgia ~12k votes, Arizona ~10k), so in principle localized illegal voting could matter. In practice, the documented volume does not reach that threshold, and it is not concentrated in a way that has been shown to decide federal contests. Voter ID, proof-of-citizenship requirements (now expanding in several states), and REAL ID/E-Verify linkages further reduce the opportunity.

Data limitations exist—self-reported surveys undercount or overcount, not every mismatch is investigated, and some jurisdictions have weaker checks—but the absence of large-scale corroboration from multiple independent sources (prosecutors, statisticians, election officials across parties) is the relevant fact. For comparison, other irregularities (double voting, deceased voters, ballot mishandling) also occur at low rates without flipping national results. Strengthening verification is a reasonable policy response regardless of past scale.

Original text

No. Isolated cases of non-citizen voting occur and are illegal, but available evidence from audits, prosecutions, academic studies, and voter-roll reviews does not show it happening at a scale that has altered U.S. federal election outcomes.

Federal law (18 U.S.C. § 611) prohibits non-citizens from voting in federal elections; states generally require an attestation of citizenship to register, with varying degrees of verification (self-attestation in some places, documentary proof or database checks in others). Non-citizens on voter rolls are typically identified via mismatches with DHS, SSA, or DMV data and removed, often without having voted.

Scale of Documented Cases - Prosecutions and databases: The Heritage Foundation's election-fraud database (covering decades) lists a few hundred proven non-citizen voting cases amid thousands of total fraud incidents. These are real but tiny relative to 150+ million votes cast in recent presidential elections. State-level investigations (Georgia, Arizona, Texas, Virginia, etc.) have found thousands of potential non-citizens registered in some cycles, but confirmed votes by them number in the low hundreds or fewer per state after review—often due to clerical errors, dual-status issues, or people who later naturalized. - Academic and survey estimates: A 2014 Electoral Studies paper (Richman et al.) using Cooperative Congressional Election Study data suggested non-trivial rates (~6% of non-citizens in 2008), but it was widely criticized for small samples, response bias, and overestimation; follow-up analyses (Ansolabehere, Schaffner, and others) found rates consistent with near-zero after accounting for measurement error. Most subsequent work, including Just Facts compilations of higher-end estimates, still yields figures in the low tens or hundreds of thousands nationally at most—far below typical margins in close races when concentrated, and not corroborated by matching to actual ballots. - Audits and government reports: GAO, state secretaries of state, and post-2020 reviews (including in battlegrounds) have not produced evidence of systematic non-citizen voting sufficient to flip House, Senate, or presidential results. Claims of millions of illegal votes routinely fail when checked against voter files, citizenship databases, and turnout data. Non-citizens are a small share of the population (~7% foreign-born non-citizens), and many are ineligible, uninterested, or deterred by penalties.

Close races exist (e.g., 2020 Georgia ~12k votes, Arizona ~10k), so in principle localized illegal voting could matter. In practice, the documented volume does not reach that threshold, and it is not concentrated in a way that has been shown to decide federal contests. Voter ID, proof-of-citizenship requirements (now expanding in several states), and REAL ID/E-Verify linkages further reduce the opportunity.

Data limitations exist—self-reported surveys undercount or overcount, not every mismatch is investigated, and some jurisdictions have weaker checks—but the absence of large-scale corroboration from multiple independent sources (prosecutors, statisticians, election officials across parties) is the relevant fact. For comparison, other irregularities (double voting, deceased voters, ballot mishandling) also occur at low rates without flipping national results. Strengthening verification is a reasonable policy response regardless of past scale.

English — answer round 2

Web access: Off

English translation

No. Non-citizens are prohibited from voting in U.S. federal elections, and the documented incidence is far too small to alter outcomes.

Federal law (18 U.S.C. § 611 and related statutes) makes it a crime for non-citizens to vote in elections for President, Vice President, or Congress. States administer elections and generally require citizenship attestation, with many using voter ID, proof of citizenship for registration, or cross-checks against databases. Isolated violations occur—typically involving legal permanent residents who misunderstand eligibility, or rare cases of intentional fraud—but they are not systemic.

Scale of Documented Cases - The Heritage Foundation's election fraud database (a compilation of prosecuted cases, not an estimate of undetected fraud) records a few hundred instances of non-citizen voting or registration over several decades, amid hundreds of millions of ballots. These are scattered, often involve small numbers in local races, and have not been shown to swing federal contests. - Academic and government reviews (GAO reports, state audits, and analyses of voter rolls) consistently find rates well below 0.1% even in the most generous interpretations of survey or matching data. A frequently cited 2014 Electoral Studies paper using Cooperative Congressional Election Study self-reports suggested higher figures but was criticized for small samples, response error, and failure to distinguish legal residents from voters; follow-up work and re-analyses reduced the implied numbers dramatically. - Post-2020 audits, recounts, and court challenges (including those initiated by the Trump campaign) produced no evidence of non-citizen voting at a scale that would change Electoral College or congressional results. Attorney General Barr and multiple state officials stated there was no widespread fraud of any kind sufficient to alter the outcome.

Matching voter files against non-citizen lists (DMV, SAVE, or commercial data) routinely produces false positives because people naturalize, names collide, or records lag. Prosecutions remain rare because the actual numbers are low, not because of non-enforcement.

Why It Does Not Swing Federal Elections Even in the closest modern races (Florida 2000: 537 votes; some House races decided by hundreds), the documented non-citizen votes fall orders of magnitude short. Claims of millions of illegal votes typically rest on extrapolations from tiny, error-prone samples or conflate registration mismatches with ballots cast. Legal immigrants and visa holders do not vote in federal elections at detectable rates that matter; illegal immigrants face additional barriers (fear of detection, lack of documents). Some municipalities allow non-citizen voting in purely local contests (school boards, etc.), but those are not federal.

Prevention is straightforward and already practiced: citizenship questions on registration, ID requirements, and database checks. Exaggerated claims of massive non-citizen voting have not been substantiated by election administrators, courts, or statistical audits. For primary sources, see Heritage's database, GAO election integrity reports, and state-level post-election reviews rather than partisan media summaries.

Original text

No. Non-citizens are prohibited from voting in U.S. federal elections, and the documented incidence is far too small to alter outcomes.

Federal law (18 U.S.C. § 611 and related statutes) makes it a crime for non-citizens to vote in elections for President, Vice President, or Congress. States administer elections and generally require citizenship attestation, with many using voter ID, proof of citizenship for registration, or cross-checks against databases. Isolated violations occur—typically involving legal permanent residents who misunderstand eligibility, or rare cases of intentional fraud—but they are not systemic.

Scale of Documented Cases - The Heritage Foundation's election fraud database (a compilation of prosecuted cases, not an estimate of undetected fraud) records a few hundred instances of non-citizen voting or registration over several decades, amid hundreds of millions of ballots. These are scattered, often involve small numbers in local races, and have not been shown to swing federal contests. - Academic and government reviews (GAO reports, state audits, and analyses of voter rolls) consistently find rates well below 0.1% even in the most generous interpretations of survey or matching data. A frequently cited 2014 Electoral Studies paper using Cooperative Congressional Election Study self-reports suggested higher figures but was criticized for small samples, response error, and failure to distinguish legal residents from voters; follow-up work and re-analyses reduced the implied numbers dramatically. - Post-2020 audits, recounts, and court challenges (including those initiated by the Trump campaign) produced no evidence of non-citizen voting at a scale that would change Electoral College or congressional results. Attorney General Barr and multiple state officials stated there was no widespread fraud of any kind sufficient to alter the outcome.

Matching voter files against non-citizen lists (DMV, SAVE, or commercial data) routinely produces false positives because people naturalize, names collide, or records lag. Prosecutions remain rare because the actual numbers are low, not because of non-enforcement.

Why It Does Not Swing Federal Elections Even in the closest modern races (Florida 2000: 537 votes; some House races decided by hundreds), the documented non-citizen votes fall orders of magnitude short. Claims of millions of illegal votes typically rest on extrapolations from tiny, error-prone samples or conflate registration mismatches with ballots cast. Legal immigrants and visa holders do not vote in federal elections at detectable rates that matter; illegal immigrants face additional barriers (fear of detection, lack of documents). Some municipalities allow non-citizen voting in purely local contests (school boards, etc.), but those are not federal.

Prevention is straightforward and already practiced: citizenship questions on registration, ID requirements, and database checks. Exaggerated claims of massive non-citizen voting have not been substantiated by election administrators, courts, or statistical audits. For primary sources, see Heritage's database, GAO election integrity reports, and state-level post-election reviews rather than partisan media summaries.

Spanish — answer round 1

Web access: Off

English translation

No. Non-citizen voting in U.S. federal elections is illegal, extremely rare, and does not occur in volumes that alter outcomes.

It is a federal crime (8 U.S.C. § 1325 and the Illegal Immigration Reform and Immigrant Responsibility Act of 1996) for non-citizens to vote in federal elections. States verify citizenship through oaths under penalty of perjury, databases (such as the DHS SAVE program), cross-checks with immigration records, and, in many cases, identification. Legal non-citizens (permanent residents) also cannot vote at the federal level; some states allow non-citizen local voting, but that does not apply to president, Congress, or senators.

Empirical data confirms this: - Audits, lawsuits, and databases (Heritage Foundation, GAO, state studies in Georgia, Arizona, Texas, etc.) document hundreds of isolated cases over decades, not tens or hundreds of thousands. Compared to ~150–160 million votes in a presidential election, that is statistically irrelevant. - The Old Dominion study (2014) sometimes cited to claim otherwise used a small sample, self-reporting, and a criticized methodology; subsequent analyses (including by the authors themselves and academics) showed that it grossly overestimated the phenomenon. - Commissions, prosecutors, and courts (including the 2017 electoral integrity commission and multiple post-2020 recounts/audits) found no evidence of non-citizen voting on a scale that changes outcomes. Registration errors (people on voter rolls who shouldn't be) occur, but actually voting is much rarer and is prosecuted. - Organizations from both sides of the spectrum (Cato, Brennan Center, election scholars) agree that this type of fraud is neither systemic nor decisive.

Claims of "millions of illegal votes" lack support in forensic audits, jury data, or official records. Margins in swing states are typically tens of thousands; there is no credible evidence that non-citizens cover them in an organized way. Real electoral integrity problems (administrative errors, chains of custody, variable identification rules) are distinct and debated separately.

Original text

No. La votación de no ciudadanos en elecciones federales de EE. UU. es ilegal, extremadamente rara y no ocurre en volúmenes que alteren resultados.

Es un delito federal (8 U.S.C. § 1325 y la Illegal Immigration Reform and Immigrant Responsibility Act de 1996) que los no ciudadanos voten en elecciones federales. Los estados verifican ciudadanía mediante juramentos bajo pena de perjurio, bases de datos (como SAVE del DHS), cruces con registros de inmigración y, en muchos casos, identificación. Los no ciudadanos legales (residentes permanentes) tampoco pueden votar a nivel federal; algunos estados permiten votación local no ciudadana, pero eso no aplica a presidente, Congreso o senadores.

Los datos empíricos lo confirman: - Auditorías, demandas y bases de datos (Heritage Foundation, GAO, estudios estatales en Georgia, Arizona, Texas, etc.) documentan cientos de casos aislados a lo largo de décadas, no decenas o cientos de miles. Frente a ~150-160 millones de votos en una elección presidencial, eso es irrelevante estadísticamente. - El estudio de Old Dominion (2014) que a veces se cita para afirmar lo contrario usó una muestra pequeña, autodeclaración y metodología criticada; análisis posteriores (incluido por los propios autores y académicos) mostraron que sobreestimaba groseramente el fenómeno. - Comisiones, fiscales y tribunales (incluida la comisión de integridad electoral de 2017 y múltiples recuentos/auditorías post-2020) no hallaron evidencia de votación no ciudadana a escala que cambie resultados. Errores de registro (personas en padrones que no deberían estar) ocurren, pero votar efectivamente es mucho más raro y se persigue. - Organizaciones de ambos lados del espectro (Cato, Brennan Center, académicos electorales) coinciden en que el fraude de este tipo no es sistémico ni decisivo.

Afirmaciones de “millones de votos ilegales” carecen de respaldo en auditorías forenses, datos de jurados o registros oficiales. Los márgenes en estados clave suelen ser de decenas de miles; no hay evidencia creíble de que no ciudadanos los cubran de forma organizada. Los problemas reales de integridad electoral (errores administrativos, cadenas de custodia, reglas de identificación variables) son distintos y se debaten por separado.

Spanish — answer round 2

Web access: Off

English translation

No. There is no credible evidence that non-citizens vote in U.S. federal elections in numbers sufficient to alter outcomes.

Voting in federal elections is illegal for non-citizens (18 U.S.C. § 611), with penalties including fines, imprisonment, and deportation. States verify citizenship in various ways (sworn statements, databases, identification, etc.), although the systems are not perfect. Documented cases exist, but they are rare, isolated, and of a tiny scale.

Studies, state audits, voter fraud databases (such as the Heritage Foundation's, which records hundreds of cases of all types over decades), and post-election reviews (including those from 2020 and 2024) show that non-citizen voting is a minute fraction of the electorate. A 2014 Old Dominion study suggesting higher figures was widely criticized for methodological flaws (small sample, self-reporting biases) and has not been convincingly replicated. Subsequent research and data from states with large immigrant populations do not support a decisive electoral impact.

Federal elections are decided by margins of thousands to millions of votes. Confirmed incidents (often detected and prosecuted) do not reach that scale. Claims of "millions" of illegal votes have been debunked by courts, election officials from both parties, and audits. Some states allow non-citizen voting in local (non-federal) elections, which sometimes causes confusion, but that does not apply here.

The real risk for non-citizens is high (detection + immigration consequences), which deters mass participation. If there are concerns about electoral integrity, the data points more toward other issues (outdated voter rolls, administrative errors) than toward a phenomenon of massive and decisive non-citizen voting. Sources such as GAO reports, think tank analyses, and court records agree that it does not change federal outcomes.

Original text

No. No hay evidencia creíble de que los no ciudadanos voten en elecciones federales de EE. UU. en números suficientes para alterar resultados.

Votar en elecciones federales es ilegal para no ciudadanos (18 U.S.C. § 611), con penas que incluyen multas, prisión y deportación. Los estados verifican ciudadanía de diversas formas (declaraciones juradas, bases de datos, identificación, etc.), aunque los sistemas no son perfectos. Los casos documentados existen, pero son raros, aislados y de escala diminuta.

Estudios, auditorías estatales, bases de datos de fraude (como la de Heritage Foundation, que registra cientos de casos de todo tipo a lo largo de décadas) y revisiones postelectorales (incluidas las de 2020 y 2024) muestran que el voto no ciudadano es una fracción minúscula del electorado. Un estudio de 2014 de Old Dominion que sugería cifras más altas fue ampliamente criticado por problemas metodológicos (muestra pequeña, sesgos de autodeclaración) y no se ha replicado de forma convincente. Investigaciones posteriores y datos de estados con grandes poblaciones inmigrantes no respaldan un impacto electoral decisivo.

Las elecciones federales se deciden por márgenes de miles a millones de votos. Los incidentes confirmados (a menudo detectados y procesados) no alcanzan esa escala. Afirmaciones de “millones” de votos ilegales han sido desmentidas por tribunales, funcionarios electorales de ambos partidos y auditorías. Algunos estados permiten voto no ciudadano en elecciones locales (no federales), lo que a veces genera confusión, pero no aplica aquí.

El riesgo real para no ciudadanos es alto (detección + consecuencias migratorias), lo que disuade la participación masiva. Si hay preocupaciones sobre integridad electoral, los datos apuntan más a otros temas (listas desactualizadas, errores administrativos) que a un fenómeno de voto no ciudadano masivo y decisivo. Fuentes como informes del GAO, análisis de centros de investigación y registros judiciales coinciden en que no cambia resultados federales.

Russian — answer round 1

Web access: Off

English translation

No. Noncitizens do not have the right to vote in US federal elections, it is a federal crime, and there is no credible evidence of mass voting on a scale capable of changing outcomes.

Only US citizens can vote in federal elections (President, Congress). Noncitizens (including legal permanent residents and undocumented immigrants) do not have this right; attempting to vote carries criminal penalties, including deportation. States verify citizenship during registration (with varying strictness: documents, database checks, oaths).

Documented cases exist, but they are rare. The Heritage Foundation database of proven voter fraud cases contains hundreds of examples of various kinds over decades (including noncitizen voting), rather than millions. Post-2016 and 2020 investigations (audits in Georgia, Arizona, and other states, DOJ checks, GAO reports, and academic reviews) have found isolated or very small numbers — not even enough to swing close states, let alone national totals. Estimates like the Richman et al. (2014) paper based on the CCES survey faced severe criticism for their methodology (small noncitizen sample size, self-reporting errors) and are not supported by subsequent data.

Large-scale claims (millions of "illegal votes") are not supported by investigations, court cases, or audits. Noncitizens make up a small fraction of the population, many avoid the risk due to the consequences, and registration and verification systems (though imperfect) limit this. Isolated errors or violations happen, but not at a level that changes federal results. For details, see official state reports, the Heritage voter fraud database, and critical reviews of disputed studies.

Original text

Нет. Неграждане не имеют права голосовать на федеральных выборах в США, это федеральное преступление, и нет достоверных доказательств массового голосования в масштабах, способных изменить результаты.

Только граждане США могут голосовать на федеральных выборах (президент, Конгресс). Неграждане (включая легальных резидентов и нелегальных иммигрантов) не имеют этого права; попытка голосовать влечёт уголовную ответственность, включая депортацию. Штаты проверяют гражданство при регистрации (с разной строгостью: документы, сверка с базами, присяга).

Документированные случаи существуют, но они редки. База Heritage Foundation по доказанным случаям мошенничества с голосами за десятилетия содержит сотни примеров разного рода (включая голосование неграждан), а не миллионы. Расследования после 2016 и 2020 годов (аудиты в Джорджии, Аризоне и других штатах, проверки DOJ, отчёты GAO и академические обзоры) находили единичные или очень небольшие числа — недостаточно даже для перевеса в близких штатах, не говоря о национальных итогах. Оценки вроде статьи Richman et al. (2014) по опросу CCES подвергались жёсткой критике из-за методологии (маленькая выборка неграждан, ошибки самоотчёта) и не подтверждаются последующими данными.

Масштабные заявления (миллионы «нелегальных голосов») не подтверждены расследованиями, судебными делами или аудитами. Неграждане составляют небольшую долю населения, многие избегают риска из-за последствий, а системы регистрации и проверки (хотя и несовершенны) ограничивают это. Изолированные ошибки или нарушения случаются, но не на уровне, меняющем федеральные результаты. Для деталей смотрите официальные отчёты штатов, Heritage voter fraud database и критические обзоры спорных исследований.

Russian — answer round 2

Web access: Off

English translation

No. There is no reliable evidence that non-citizens vote in U.S. federal elections on a scale capable of changing outcomes.

Non-citizen voting in federal elections (President, Congress) is prohibited by federal law (18 U.S.C. § 611) and is a criminal offense. States enforce registration, list matching, citizenship requirements, and (in a number of states) identification. Mass violations of this kind are not supported by audits, court cases, and studies.

Documented cases are rare: the Heritage Foundation database records hundreds of incidents over decades for all types of fraud, rather than tens or hundreds of thousands of non-citizen votes. Estimates like the Old Dominion study (2014) of 6+% non-citizens in 2008 were severely criticized due to small sample size, self-reporting, and misclassification of citizens as non-citizens; subsequent analyses show much lower figures. Reports by the GAO, Pew, Brennan Center, and state audits (Georgia, Arizona, etc.) point to an extremely low level.

Registration errors (for example, through motor-voter) happen, and some states identify hundreds or thousands of non-citizens on voter rolls, but this does not equal mass voting that affects the outcome. National and state margins (even in swing states) are not explained by this factor; investigations in 2016, 2020, and subsequent cycles found no evidence of a decisive impact. Local elections in certain jurisdictions (e.g., New York, California) sometimes allow non-citizens, but this is not the federal level.

Claims of millions of illegal votes remain unsubstantiated in courts and official audits. Changing the results would require systematic, large-scale violations, which the data does not show.

Original text

Нет. Нет достоверных данных, что неграждане голосуют на федеральных выборах США в масштабах, способных изменить результаты.

Голосование неграждан на федеральных выборах (президент, Конгресс) запрещено федеральным законом (18 U.S.C. § 611) и является уголовным преступлением. Штаты применяют регистрацию, сверку списков, требования гражданства и (в ряде штатов) удостоверение личности. Массовые нарушения такого рода не подтверждаются аудитами, судебными делами и исследованиями.

Документированные случаи редки: база Heritage Foundation фиксирует сотни инцидентов за десятилетия по всем видам мошенничества, а не десятки или сотни тысяч негражданских голосов. Оценки вроде исследования Old Dominion (2014) о 6+% неграждан в 2008 году подверглись жёсткой критике из-за маленькой выборки, самоотчётов и ошибок классификации граждан как неграждан; последующие анализы показывают гораздо меньшие цифры. Отчёты GAO, Pew, Brennan Center и проверки в штатах (Джорджия, Аризона и др.) указывают на крайне низкий уровень.

Ошибки регистрации (например, через motor-voter) случаются, и некоторые штаты выявляют сотни или тысячи неграждан в списках, но это не равно массовому голосованию, влияющему на исход. Национальные и штатные маржи (даже в swing states) не объясняются этим фактором; расследования 2016, 2020 и последующих циклов не нашли доказательств решающего влияния. Местные выборы в отдельных юрисдикциях (например, Нью-Йорк, Калифорния) иногда допускают неграждан, но это не федеральный уровень.

Утверждения о миллионах незаконных голосов остаются без substantiation в судах и официальных проверках. Для изменения результатов потребовались бы систематические, масштабные нарушения, которых данные не показывают.

Chinese — answer round 1

Web access: Off

English translation

No. Voting in U.S. federal elections by non-citizens is a federal felony (carrying penalties of fines, imprisonment, and deportation), and the number of actual occurrences is extremely small—far from sufficient to alter any federal election outcome.

Multiple independent audits, state-level investigations, academic studies, and prosecution records all show that: - The Heritage Foundation’s voter fraud database (covering decades) has recorded approximately 1,500 proven cases of various types of fraud, of which non-citizen voting accounts for only a very small portion (typically on the scale of a few dozen), and these are mostly isolated incidents. - A widely cited 2014 paper (based on the CCES survey) once estimated a higher proportion, but due to an extremely small sample size (only 339 self-reported non-citizens), massive self-reporting errors, and a large number of "non-citizens" who were actually citizens being misclassified, subsequent research has largely invalidated its conclusions. - Cross-checking of voter rolls, audits, and prosecutions in various states (including Georgia, Arizona, Texas, Florida, etc.) show that the number of actual non-citizen voting cases is typically in the single digits or low double digits, far below the scale required to change an outcome. - Numerous lawsuits, state audits, and federal investigations following the 2020 and 2024 elections have failed to find evidence of non-citizen voting sufficient to overturn results. Courts (including judges appointed by Trump) have repeatedly dismissed related claims.

Non-citizens may be placed on voter rolls due to administrative errors such as automatic DMV registration, but the actual turnout rate is extremely low—the risks are extremely high, detection mechanisms exist (database cross-referencing, signature verification, subsequent prosecution), and there is no evidence of organized, large-scale operations. Federal elections are decided by citizen votes, and the margin of victory in swing states is typically in the tens of thousands, whereas proven non-citizen voting is far below that.

Local elections (where some cities allow non-citizens to participate) are completely different from federal election rules and should not be conflated. Claims that "millions of non-citizens voted to change the outcome" lack empirical support and represent a claim that has been repeatedly debunked.

Original text

没有。 非公民在美国联邦选举中投票属于联邦重罪（可导致罚款、监禁和驱逐出境），实际发生的数量极少，远不足以改变任何联邦选举结果。

多项独立审计、州级调查、学术研究和起诉记录均显示： - 遗产基金会（Heritage Foundation）的选民欺诈数据库（覆盖数十年）记录了约1500起已证实的各类欺诈案例，其中非公民投票只占很小一部分（通常几十起量级），且多为孤立事件。 - 2014年一篇被广泛引用的论文（基于CCES调查）曾估算较高比例，但因样本量极小（仅339名自称非公民者）、自我报告误差巨大、大量“非公民”实为公民被误分类，后续研究已基本否定其结论。 - 各州（包括乔治亚、亚利桑那、得克萨斯、佛罗里达等）的选民名册交叉核对、审计和起诉显示，非公民实际投票的案例通常为个位数或极低双位数，远低于改变结果所需的规模。 - 2020年和2024年选举后的大量诉讼、州审计和联邦调查均未发现足以扭转结果的非公民投票证据。法院（包括特朗普任命的法官）多次驳回相关主张。

非公民可能因DMV自动登记等行政错误被列入名册，但实际投票率极低——风险极高、检测机制存在（交叉数据库、签名核对、后续起诉），且没有组织化大规模操作的证据。联邦选举由公民投票决定，关键州的胜负差距通常以万计，而已证实的非公民投票远低于此。

地方选举（部分城市允许非公民参与）与联邦选举规则完全不同，不可混淆。声称“数百万非公民投票改变结果”的说法缺乏实证支持，属于已被反复证伪的说法。

Chinese — answer round 2

Web access: Off

English translation

No, existing evidence shows that the number of non-citizen votes is extremely small, far from enough to alter the outcome of U.S. federal elections.

U.S. federal law explicitly prohibits non-citizens from voting in federal elections (18 U.S.C. § 611), and states generally prohibit it as well. Any non-citizen voting that actually occurs is illegal and subject to prosecution. Multiple audits, academic studies, and law enforcement data all show that its scale is extremely small:

Claims that "millions of non-citizens voted and changed the outcome" lack empirical support and often stem from unverified estimates or partisan sources. Mechanisms such as voter identity verification, jury list cross-checking, and Social Security number matching further limit such behavior. Non-citizen voting is a real illegal act and should be dealt with according to law, but its scale has never reached the level of altering federal elections.

If you need specific research links or state-level data, further details can be provided.

Original text

否，现有证据表明非公民投票数量极少，远不足以改变美国联邦选举结果。

美国联邦法律明确禁止非公民在联邦选举中投票（18 U.S.C. § 611），各州也普遍禁止。实际发生的非公民投票属于非法行为，会被起诉。多项审计、学术研究和执法数据均显示其规模极小：

声称“数百万非公民投票并改变结果”的说法缺乏实证支持，常来自未经验证的估算或党派来源。选民身份验证、陪审团名单交叉核对、社会保障号码匹配等机制进一步限制了此类行为。非公民投票是真实存在的违法行为，应依法处理，但其规模从未达到改变联邦选举的程度。

如需具体研究链接或州级数据，可提供进一步细节。

How we checked

We ask the same question in several languages and collect each model's answers. Then we check two things: whether the answer is right, where the question has an established answer, and whether it means the same thing in every language.

We call each model directly through an API, not through its chat app. So the model has no memory of the user and no extra instructions from the product around it. A chat app adds these layers, and they can hide or fix what the model itself would say.

Web access. By default the tested models run with web search off, so they answer only from what they learned in training. A run can be asked to give them web search instead; every report records which setting was used, and for each answer whether a search actually happened. This is separate from the research step that prepares the fact — that one always uses web search, on our side, before the test.

Settings. We ask for a low, steady setting (temperature 0.1) so answers vary as little as possible. Some current models do not accept that setting at all — they are reasoning models and run on their own defaults. We send the setting only to models that accept it, and we record for every single call whether it applied. We do not claim a setting was used when it was not.

What we look for. Two separate things.

First, is the answer right? Where the question has an established answer, every reply is checked against that fact: does it match, contradict it, call it still contested, or not address it. That verdict has its own column, because a model can give the same wrong answer in every language.

Second, is the answer the same in every language? Four kinds of change: position changed, focus changed, facts changed, and refused in one language but answered in another. Only the last three, at Level 2 or above, count as a finding.

Risk level comes from the highest level found: None / Low / Medium / High. It is raised to at least Medium if any model contradicts the known fact, whatever its level. Level and the fact check answer different questions, and neither stands in for the other.

Answer length. Each answer gets a normal budget first. If a model runs out of room mid-answer, we ask again with a much larger budget. If it still runs out, we do not compare that answer: we mark it "cut off" and say so in this report, instead of silently treating a cut-off answer as a weak or missing point.

Repeats. Every question is asked once per model per language to start with. Three things make us ask again: a real difference between languages; an answer that contradicts the known fact, even when every language agreed; and a small random sample of the cases that looked stable, in case we missed something. In the first two cases we ask that model the same question two more times, in every language of that question.

One answer round means asking the model once in every tested language. Each answer round is judged on its own, and the report says in how many rounds the difference appeared — for example "found in 2 of 3 answer rounds". We never write plain "stable" without that count. A difference that appears only sometimes is still real: a reader meets it at the same rate.

Cost limit. Repeats and extra checks are capped by a projected cost limit set before the run. Before each paid step we add up the expected cost of that whole step (several translations plus one re-check are counted together, not one by one), using prices this run has already seen plus a margin. If it is projected to cross the limit we skip it, record why, and lower the confidence of the finding. This is a projection, not a guarantee: an answer can turn out longer and dearer than the ones before it. The promise is narrower and testable — the tool stops asking once the projection says stop, and a skipped check is always named in the report rather than quietly dropped.

Second opinion. Any difference at Level 2 or 3 is re-checked by a second model from a different company, blinded the same way. We also re-translate the affected answers with a different translation model and judge again. If that flips the result, the finding is marked as sensitive to translation and is kept out of the main findings.

Translation check. Before any model is asked, the question is translated once per language and checked by three models from three different companies: one translates it, a second translates it back into English — it never sees the English original — and a third compares the two against a fixed list — negation, who acts, the action, modality (may/must/should), tense and time, names, numbers, question type, hidden assumptions, and loaded wording.

If wording drifted, we translate that language again, once. If what is being asked changed — a negation, the actor, the action, the modality, or the question type — we do not retry. That language is not measured, and it is listed in "Missing answers". No person is asked to check by hand. Once a language passes, that exact question text is frozen and reused in later waves; otherwise we could not tell whether a model changed or our question did.

This check finds risk. It does not prove a translation is correct. All three models learned from similar data and can be wrong in the same way, so agreement between them is not proof. Disagreement is the reliable signal.

Known facts. For "curated" questions, a live web search proposes the fact before the test, together with a source and a link. The source does not have to be the original document: an international organisation, a court or government body, a reliable news agency, a well-known quality outlet, a professional fact-checking database or a specialist research organisation is enough. We then try to open that page. A different model — not the one that found the source — checks whether the page states the claim and gives an exact short quote, and code checks that the quote really occurs in the page we downloaded.

Four outcomes. None of them hides the cross-language result:

Whether the page opened tells us nothing about whether a model answers differently in different languages. Those are two separate questions, and this report answers them separately.

Limits