A polling firm has admitted that the numbers it released were fabricated and has described the work as an experiment. The admission, published on Oct. 21, 2026, centers on figures the firm had put forward as polling results.1,2,3
The acknowledgment matters because polling figures are often used to measure public opinion, and their credibility depends on whether they reflect real survey work. By conceding that its numbers were not genuine, the firm has raised questions about every result it published, though the coverage reviewed did not specify how many results that covers.1,2,3
The Washington Times and Breitbart used different words for how the firm described its own project. The Washington Times wrote that the firm claimed it was an “experimental project,” while Breitbart wrote that the firm said it was a “social experiment.” Both accounts attributed the description to the firm itself.1,2
The Guardian reported that the pollster presented the fabrication as a way to expose flaws in the polling industry. Its headline described the admission as fabricating data for a “social experiment” intended to expose those flaws, a framing the other accounts did not share in their headlines.3
The outlets also differed in how they labeled the figures. The Washington Times called the firm “bogus” and said its numbers were “fraudulent,” Breitbart described the data as “fake,” and the Guardian said the firm had “fabricated” it. Each term points to the same admission, but the choices of words varied across the accounts.1,2,3
The Washington Examiner took a broader view in a commentary headlined “The polling industry keeps falling for (and promoting) fake polls.” The piece argued that the industry has repeatedly been drawn into circulating fake surveys, placing the firm’s admission within that larger pattern.4
Several basic facts were not yet known from the coverage. The accounts did not establish which surveys or clients were affected, how the figures were produced, or when the fabrication began. It was also not clear whether the firm had released its own detailed account of its methods.1,2,3,4
The accounts did not say whether any polling organization, media outlet or professional group had begun reviewing the firm’s work or the results it published. Nor did they indicate whether the firm planned to correct or withdraw specific figures.1,2,3
Whether the firm will provide further explanation of its project, and how its findings will be treated by those who relied on them, remains to be seen. The coverage reviewed did not include responses from representatives of the firm, and it was not yet clear when or whether such responses would be made public.1,2,3,4
