The fake poll had reversed the actual standing entirely
As Peru approaches its June 7 presidential runoff, a fabricated poll attributed to the respected Instituto de Estudios Peruanos began circulating on Facebook, falsely portraying Keiko Fujimori leading Roberto Sánchez by two points — a reversal of the actual standings recorded in IEP's genuine April survey. The institute, which had conducted no polls in May, moved swiftly to disavow the forgery. In the fragile weeks before a consequential vote, the episode reminds us that electoral perception itself has become a contested terrain, where invented numbers can quietly reshape the story a democracy tells itself.
- A forged poll dressed in IEP's visual identity spread across Facebook, claiming Fujimori led Sánchez 39–37% — numbers that inverted the real race entirely.
- The fabrication was not casual: someone carefully replicated IEP's color scheme and bar graphics while altering candidate photos, sample sizes, and response categories to manufacture credibility.
- Fact-checkers found the seams — a misplaced logo, a reshaped graphic, a phantom response option, and a sample figure that matched no published IEP methodology.
- IEP confirmed through its official channels that it had published exactly one poll in April and had conducted zero surveys in May, cutting the forgery off at its source.
- With the June 7 runoff approaching and voters still undecided, the circulation of false polling data poses a direct risk of distorting public perception of which candidate holds genuine momentum.
In mid-May, a poll began spreading on Facebook purporting to show Keiko Fujimori holding a two-point lead over Roberto Sánchez ahead of Peru's June 7 presidential runoff. The image bore the branding of the Instituto de Estudios Peruanos — Peru's well-regarded polling institute — and presented figures with the visual confidence of an official publication: 39 percent for Fujimori, 37 percent for Sánchez.
The problem was that IEP had published no such survey. Its most recent authentic poll, conducted in late April, told a different story: Sánchez led with 32 percent to Fujimori's 31 percent. The fake poll had not merely invented a Fujimori advantage — it had reversed the actual standings, inflating her support by eight points and understating Sánchez's by five.
The distortions ran deeper than the headline numbers. The real IEP survey recorded 13 percent undecided voters and 24 percent blank or spoiled ballots; the forgery compressed those figures to 9 and 12 percent respectively, and added a response category — 'none/would not vote' — that IEP's methodology does not use. The forger had borrowed IEP's characteristic green and orange candidate bars and overall layout, but left behind revealing errors: a resized graphic circle, candidates photographed in the wrong order, a sample size of 1,208 rather than IEP's actual 1,197, and the IEP logo placed at the bottom right rather than at the top alongside La República's branding, as authentic surveys display it.
IEP responded through its official X account, stating plainly that it had conducted one poll in April and no surveys in May. The clarification was both a correction and a signal — false electoral data was moving through Peru's information environment at the precise moment voters were forming their final judgments about who leads and who trails in one of the country's most consequential political contests.
A fabricated poll began circulating on Facebook in mid-May, claiming to show Keiko Fujimori leading Roberto Sánchez by two percentage points in Peru's upcoming presidential runoff. According to the fake survey, Fujimori held 39 percent support while Sánchez commanded 37 percent—a narrow advantage that could shape perceptions of momentum heading into the June 7 vote. The problem was simple: the Instituto de Estudios Peruanos, Peru's respected polling institute, had never published any such survey.
When fact-checkers examined the claim against IEP's actual public record, the discrepancies became immediately apparent. The institute's most recent authentic poll dated to late April, not mid-May as the viral image suggested. More importantly, that real survey told a different story entirely. Sánchez, running under the Juntos por el Perú banner, registered 32 percent support, while Fujimori of Fuerza Popular trailed at 31 percent. The fake poll had not only invented a lead for Fujimori but reversed the actual standing, inflating her numbers by eight points while understating Sánchez's by five.
The fabrication extended beyond the headline figures. The authentic IEP survey showed 13 percent of respondents remained undecided, yet the fake version claimed only 9 percent fell into that category. Blank and spoiled ballots registered at 24 percent in the real poll but appeared as 12 percent in the counterfeit. The forged image also included a response option—"none/would not vote"—that did not exist in IEP's actual methodology. Someone had constructed this false poll with deliberate attention to detail, borrowing visual elements from genuine IEP surveys: the characteristic green and orange bars representing each candidate, the overall layout and color scheme.
Yet the forgery contained telltale signs of its inauthenticity for anyone examining it closely. The circle in the main graphic had been resized and reshaped in ways inconsistent with IEP's standard design. The photographs of the two candidates appeared in a different order than IEP typically used. The technical specifications claimed the survey had sampled 1,208 respondents, while the actual April poll specified 1,197 people who said they would vote. Most glaring, the IEP logo appeared in the lower right corner of the fake poll, when authentic surveys always positioned it at the top alongside the logo of La República, the news organization with which IEP maintains a partnership.
The institute itself moved quickly to address the deception. Through a statement posted on its official X account, IEP made the situation unambiguous: the organization had published one poll in April and had conducted no surveys in May whatsoever. The statement served as both a correction and a warning, signaling that false electoral data was circulating in the information ecosystem as Peru approached one of its most consequential political moments. With the runoff scheduled for early June and voters still forming their final judgments, the appearance of fabricated polling data—especially data that reversed the actual race dynamics—represented a direct attempt to manipulate public perception of the candidates' relative strength.
Citas Notables
The Instituto de Estudios Peruanos only published one poll in April and has not conducted or released any survey in May— IEP official statement via X
La Conversación del Hearth Otra perspectiva de la historia
Why would someone go to the trouble of forging a poll this detailed? It's not like they just made up random numbers.
Right. Whoever created this understood IEP's actual design, their methodology, their partnerships. They were trying to pass as legitimate, not just throw something together. That takes intent and knowledge.
And the numbers they chose—39 to 37—that's close enough to seem plausible but gives Fujimori the edge. Was that calculated?
Almost certainly. A ten-point spread would look absurd. But two points? That's the kind of margin that sticks in people's minds, that makes them think momentum is shifting. It's designed to feel real.
The actual poll showed Sánchez ahead by one point. So this fake poll didn't just lie—it flipped the entire narrative.
Exactly. It took the real race, which was essentially tied, and created a false story of Fujimori gaining ground. That's not sloppy misinformation. That's strategic.
How many people do you think saw the fake version before IEP corrected it?
We don't know. Facebook doesn't track that. But it was circulating enough that fact-checkers caught it and IEP felt compelled to issue a public statement. That suggests it had real reach.
What happens now? Does the correction undo the damage?
That's the hard part. Studies show corrections often don't fully erase false beliefs. Some people saw the fake poll and will remember it as real, even after the correction. The lie gets out faster than the truth.