Viral Claim of 228,000 Fake Mamdani Votes Debunked by Election Experts

A viral claim that a quarter-million votes for New York City Mayor Zohran Mamdani were fraudulent has been circulating online for weeks, but the audit behind it never actually tied a single vote to Mamdani, and its core methodology rests on a misunderstanding of how Social Security numbers work. The claim traces back to a report released in late July by a group called New York Citizens Audit Civic Fund, which examined a small sample of voter registrations pulled from a commercial database rather than official state voter rolls.

According to a Reuters fact check, the group's Real America Vote analysis looked at 10,000 registered voters selected from a subset of people who registered between 2021 and 2025, and estimated that roughly one quarter of that sample were unqualified voters — about 2,480 people. From that narrow sample, the audit extrapolated a claim that as many as one million unqualified voters could be on New York City's rolls, and that more than 200,000 suspicious votes may have been cast citywide. Social media posts amplifying the findings claimed fraudulent votes were cast by people without valid Social Security numbers, by non-existent voters, and by voters located outside New York City.

Where the Social Security Math Falls Apart

The audit flagged voter records with Social Security numbers beginning with 700 and 800, assuming those numbers belonged only to people born after 2011. But that assumption is outdated. The Social Security Administration switched to fully randomized number assignment on June 25, 2011, a policy change that made numbers in the 700s and 800s available to any new applicant regardless of age, according to AARP. That means adult immigrants and naturalized citizens who applied for a Social Security number after mid-2011 could legitimately hold numbers in those ranges, not just children.

Charles Stewart III, a political science professor and director of the MIT Election Data and Science Lab, told Reuters that Social Security numbers beginning with 700 and 800 could belong to adults, undercutting the audit's central assumption. Stewart also said the margin of error the group cited covers only random sampling error and does not establish whether the underlying commercial database correctly classified voters in the first place. That distinction matters because the entire extrapolation to nearly a million supposedly unqualified voters and more than 200,000 suspicious votes rests on the accuracy of that classification.

Commercial Data Versus Official Voter Rolls

The audit relied on a commercial database whose accuracy was never independently verified, rather than pulling directly from New York City Board of Elections records. Stewart said such commercial databases tend to underrepresent younger, minority, and residentially mobile populations, which can generate false-positive errors when used to flag supposedly invalid voters. Research from the MIT Election Data and Science Lab, cited by Reuters, found that these consumer marketing lists lean heavily on credit histories and change-of-address filings that often lag behind where people actually live.

New York State's actual voter registration process requires applicants to provide either a state driver's license number or the last four digits of a Social Security number, both of which are cross-checked against state DMV and federal databases before registration is approved, per the New York State Board of Elections. First-time voters whose identities can't be verified in advance still must show photo ID or proof of address at the polls. New York also already runs a legally mandated post-election check: under state election law, county boards must manually audit paper records from at least 3 percent of voting machines within 15 days of a general election, according to the New York State Senate.

No Votes Actually Tied to Mamdani

Despite headlines suggesting otherwise, the Real America Vote report never tied specific votes to Mamdani and did not find 228,000 invalid Mamdani votes or 228,000 unique unqualified voters, Reuters reported. A conservative influencer claimed in an August 2 post on X that an independent election auditor had found more than 200,000 suspicious votes in the mayoral race, and Rudy Giuliani cited the audit on a talk show, but neither claim matches what the underlying report actually says. Mamdani won the November 4, 2025 election with more than 50 percent of the vote in a contest that drew more than 2 million voters citywide.

The New York City Board of Elections told Reuters that the report used flawed methodology and inaccurate assumptions about its data. The board said it will review every submission that includes specific names, addresses, and supporting evidence, and that it has referred such matters to law enforcement agencies on multiple occasions in the past when evidence of criminal activity or election fraud surfaces. Reuters noted that the Mayor's Office referred questions to the Board of Elections, since the mayor's office does not oversee municipal elections. Both Real America Vote and the U.S. Social Security Administration did not respond to requests for comment from Reuters.

A Group With a History of Election Complaints

This is not the first time New York Citizens Audit has drawn scrutiny from state officials. New York Attorney General Letitia James issued a cease-and-desist letter in September 2023 ordering the group to stop door-to-door voter canvassing, after complaints that volunteers confronted residents at home while falsely claiming to be board of elections officials, according to the Associated Press. State officials at the time investigated reports of similar door-to-door activity across 13 New York counties.

More recently, the New York Citizens Audit Civic Fund filed a federal lawsuit in October 2025 against James and state election officials, seeking to block future election certifications until an independent audit of the state's voting system is completed, per Democracy Docket. A federal court dismissed the majority of the group's claims this month. Separately, federal law under the National Voter Registration Act of 1993 bars election administrators from conducting systematic voter roll purges within 90 days of a federal election, which is part of why inactive voters or outdated addresses can legally remain on the rolls without indicating fraud.

The fraud claim arrives amid a broader pattern of digital misinformation aimed at Mamdani's administration since he took office. In February, Mamdani publicly called on municipal, state, and federal governments to strengthen regulations against digital disinformation after AI-generated images falsely linking him to Jeffrey Epstein spread widely online shortly after his inauguration.