The company described itself as an experimental effort intended to demonstrate how a made-up polling firm could sway public opinion while presenting itself as trustworthy. The admission came as the independent polling watchdog FiftyPlusOne prepared to publish an investigation alleging Wedgewood had invented its numbers.

Before the disclosure, Wedgewood’s polls had reached audiences beyond its own social media account. They were included in several polling aggregators, including The New York Times’ collection, and appeared in two editions of POLITICO’s Morning Score polling roundup. Some candidates have said the false numbers hurt their campaigns, but the available details do not establish which campaigns were affected or whether the polls changed voters’ choices.

Wedgewood had raised thousands of dollars through crowdfunding, but its backers, operators and motives remained unclear. Possible explanations raised publicly included an effort to influence Democratic primaries, manipulate betting markets or affect campaign spending. There was no confirmed evidence identifying who was behind the project or establishing that any of those theories was true.

The company’s disclosure underscores how quickly questionable political data can acquire credibility when it is repeated by accounts and services that audiences rely on. For voters and campaigns, polling can shape expectations about which candidates are viable and where resources should go; fabricated numbers risk distorting those judgments even when their direct effect on an election cannot be measured.

By Wednesday, earlier poll PDFs associated with Wedgewood were no longer available online, and an email sent to the address listed on the company’s X account bounced back. The scale and consequences of the deception remained unresolved.