A sham polling firm admitted it fabricated surveys that upended a Michigan Senate race and fooled major aggregators, calling the stunt far too easy.
Wedgewood Polls publicly confessed to pumping out fake numbers after election analysts at FiftyPlusOne caught the fabrication and prepared to publish an investigation, the New York Post reported.
The outfit framed the scheme as a “social experiment.” It said the point was to test how easily a made-up firm could win traction, land in big averages, and get cited by major news organizations. One of those polls helped clear a path in Michigan’s Democratic Senate primary by hammering state Sen. Mallory McMorrow and leaving Abdul El-Sayed and Rep. Haley Stevens in a tighter field.
McMorrow later blamed a summer Wedgewood survey that put her support in the single digits for drying up money, crushing volunteer morale, and pushing her out of the race.
In a statement posted to X, the group cast itself as industry veterans who watched trust in polling collapse under weak methods. Then it spelled out the test.
Wedgewood Polls stated:
"We wanted to test a simple hypothesis: How easy is it for an entirely fabricated polling firm to gain traction, be taken seriously, and expose how little scrutinization exists in media and public commentary?"
Its own verdict followed in three words: “Far too easy.”
The firm claimed it never meant to sway an election or permanently mislead the public. It called the project a “wake-up call” and said standards for vetting, aggregating, and reporting poll data must change if a fictional shop can slip into major averages. Wedgewood boasted it had tricked key aggregators, including The New York Times, FiftyPlusOne, and VoteHub. It also claimed one legitimate poll in Nebraska while admitting the rest of the run was fake.
That is not a harmless prank when campaigns, donors, and news desks treat every new number like scripture. It is a stress test the press and the polling industrial complex failed in public view, the same ecosystem already strained by Democratic midterm polling jitters heading into Election Day.
Over the summer, Wedgewood released a Michigan survey that showed McMorrow’s support imploding into single digits. She said that reading clashed with her campaign’s internal numbers. At the time, the Democratic primary was a three-way fight among McMorrow, Abdul El-Sayed, and Rep. Haley Stevens.
After the hit poll, McMorrow said donations dried up. Volunteers were devastated. Reporters kept asking whether she would quit. She did.
On October 21, 2026, she posted on X about the damage from the late-June survey and the wider stunt. She did not say much about the human cost.
Mallory McMorrow said:
"Polling is broken. But tens of thousands of people poured their hearts, time, energy, hard-earned money into something they believed in, for a year-and-a-half, and it was treated like a game."
She added that volunteers gave their all, and that she felt she had let so many of them down. “Sure, a point was proven,” she wrote. The point came at the expense of people who cut checks and walked precincts while a fake shop treated their race like a lab rat.
When a fabricated poll helps shove a candidate out, the remaining field changes. El-Sayed and Stevens no longer faced the same three-way math. Voters who rely on honest signals got noise instead. That is how thin “research” becomes a political weapon even when the authors later plead social-science cosplay.
The Michigan mess was not the only race touched by Wedgewood’s product. A recent survey from the firm claimed Mississippi Democrat Scott Colom was dead even with incumbent Republican Sen. Cindy Hyde-Smith.
This week, Democrats’ Senate Majority PAC announced another $1.2 million behind Colom. Spokesperson Lauren French told the Post the super PAC “absolutely” did not rely on that poll for the decision.
Lauren French said:
"We don’t make spending or strategic decisions based on external polling."
Campaigns can deny any single survey drove a buy. The larger problem remains: fake data still entered the bloodstream of coverage and conversation around a live Senate race. In a cycle when parties are already shopping for closing arguments and momentum stories, junk inputs make every spend and every headline harder to trust, including the kind of messaging fights seen when Democrats bet on AI fear as a midterm pitch.
Zachary Donnini, VoteHub’s head of data science, said he contacted Wedgewood when the firm first appeared. He wanted to learn who they were.
They spoke by phone for about 15 minutes. The man Donnini took to be the head pollster said he was a Boston University college student who had previously worked with SoCal Strategies.
Donnini said he tries to give young people in that spot the benefit of the doubt. He had been a college student following politics as a hobby years earlier. After the confession, his tone shifted.
Zachary Donnini said:
"It’s disappointing to see how this has unfolded, because it’s going to be a h*** of a lot harder for the next college kid."
That is the predictable hangover. Gatekeepers who waved through a thin origin story now face pressure to slam the door on everyone. Serious newcomers pay for the fraud. The public pays twice, once when the fake numbers circulate, and again when skepticism hardens into cynicism.
Federal Communications Commission Chairman Brendan Carr answered the revelations with a warning aimed at licensed stations. He said the FCC takes the issues seriously and reminded broadcasters what their licenses already forbid.
Brendan Carr said:
"Broadcasters are prohibited by the terms of their federal licenses from engaging in News Distortion or promoting Broadcast Hoaxes."
Carr added that broadcasters must comply with their public interest obligations. He did not list specific stations in the available remarks. The message was still plain: airing or hyping unverified political surveys is not a cost-free habit when federal licenses ban distortion and hoaxes.
Election season already strains newsrooms hungry for horse-race copy. A chairman willing to say the quiet part aloud is a start. Enforcement and basic desk skepticism still have to catch up before the next fabricated “pollster” tests the same weak seams, especially while Washington fights over core election-integrity questions such as noncitizens found on voter rolls and related reform demands.
Wedgewood is not a one-off glitch. Months earlier, Median Strategies got caught faking surveys tied to the Los Angeles mayoral primary and the Wisconsin gubernatorial primary. In August, Median claimed it had been running a “short-term social experiment.”
Two shops. Same excuse. Real primaries used as the stage.
When fabricators can mint a brand, mimic the jargon, and watch aggregators and reporters bite, the failure is institutional. Newsrooms and average-builders are supposed to ask who fielded the sample, how the respondents were reached, and whether the firm has any track record beyond a sleek graphics package. Here, a small group of self-described industry veterans showed how little of that homework got done.
Voters do not need another lecture about “trusting the science” of politics. They need fewer laundered numbers and more adults willing to withhold a headline until the pollster is real. That standard matters as much as any convention speech that frames the midterms as a stark choice for voters in a hard-fought cycle.
Wedgewood said it wanted a wake-up call. McMorrow’s donors and volunteers got the bill. Michigan’s primary map shifted after a single-digits hit job that the authors later shrugged off as a test. Mississippi saw a dead-heat claim land in the same polluted information stream. Aggregators that prided themselves on sorting signal from noise got named as marks.
Honest elections need honest inputs, ballots counted under law, rolls maintained with care, and political “data” that is not fan fiction with a margin of error. The cure is not another panel about civility in polling. It is ruthless verification before the next fake firm discovers, again, that the bar sits on the floor.
If a handful of operators can bend coverage and rattle a Senate race with spreadsheet fiction, the scandal is not only the troll. It is every gatekeeper who made the con look respectable.
