21‑Year‑Old Behind Fake Election Polls Says Experiment Went Viral

A 21‑year‑old recent college graduate operated an anonymous website that published fabricated election polls in Wisconsin, Nevada and California. The site, called Median Strategies, released numbers that showed certain candidates with leads that were not supported by any verified survey. Those figures were picked up by local officials and shared on social media before the site disappeared after questions from a Los Angeles Times reporter. In an unsigned statement posted to the site, Median Strategies described the effort as a short‑term social experiment to test how easily false polling could enter the political information ecosystem without independent verification. Rahil Prakash, the graduate who ran the site, told the Guardian he wanted to see if fabricated polls could really penetrate that ecosystem easily. He said he was surprised when the data spread as quickly as it did. One of the polls published by Median Strategies showed Los Angeles Mayor Karen Bass ahead of her election challenger by more than ten percentage points, and she publicly cited the figure. The California Post and a widely followed polling aggregator on X also shared the Nevada poll produced by the site. Credible polling organizations did not include Median’s numbers in their aggregates because they could not verify the methodology behind the surveys. Prakash said he did not expect the California political press to pick up the data on its own, nor did he anticipate the speed at which the polls circulated. He apologized to the campaigns of Karen Bass, Nithya Raman and Francesca Hong for the confusion caused. Raman is running against Bass in the November Los Angeles mayoral election, while Hong lost the Democratic primary for Wisconsin governor earlier this month after a Median poll suggested she had led the race by more than twenty points. The site also released a fake poll showing Nevada Governor Joe Lombardo ahead by five points before the state’s general election. Prakash traced his motivation to a polling miss in the Michigan Senate primary, where early surveys had overstated Abdul El‑Sayed’s lead over Haley Stevens before election day. He said the discrepancy inspired him to test his theory about how fast misleading information can travel through media channels. As of Thursday morning the Median Strategies account on X had about twenty followers and had begun posting on 9 August, according to the Los Angeles Times. Prakash said the episode shows that a polished presentation makes false data travel easily and warned that a more resourced operation could cause far greater harm to public discourse. He said he built the website alone using artificial intelligence tools and did not receive any financial compensation for the project. The Guardian identified Prakash by tracing domain registration data tied to the Median Strategies website; reverse‑searching the two nameservers used by the site revealed about 115 other domains sharing them, only one of which matched both the nameservers and the registrar used by Median Strategies. That matching site now displays only a GitHub 404 page, but additional web records connect it to a GitHub profile containing a file with Prakash’s name. The anonymous statement on Median’s site stressed that no individual involved received compensation or financial benefit and that the project was not funded, commissioned or directed by any campaign, party, media outlet or outside entity. When asked whether he had traded on prediction markets after publishing the sham data, Prakash said he had not. On Wednesday, before the Guardian contacted him, the American Association of Political Consultants condemned the effort, describing the publication of fabricated survey data as a deception aimed at voters, the press and campaigns that rely on public data for decision‑making. Aaron Zolla contributed digital forensics that helped trace the site’s infrastructure.