Median Strategies says two recent polls were faked, raising new questions about polling transparency

Median Strategies, a polling company that released results over the past several weeks, has acknowledged that at least two of its published surveys were faked, according to a report by The Hill published on August 22, 2026. The disclosure, made earlier in the week, adds a new episode to an already strained polling environment focused on public trust and the reliability of election-related data.

The report said the company’s faked polling involved two separate races: the Wisconsin Democratic gubernatorial primary and the Los Angeles mayoral election. The company’s announcement, as described by The Hill, did not frame the issue as a minor methodology error, but instead as a fabricated set of results, which can complicate how campaigns, media organizations, and voters interpret polling information.

The Hill described the development as “shocking,” noting that it comes as the polling industry has been trying to improve confidence in its work. In recent election cycles, polling firms and other election data providers have faced intense scrutiny over sample quality, disclosure practices, and how results are presented to the public. Fabricated figures, if confirmed, directly undermine those efforts by undercutting the reliability of a specific firm’s released outputs.

While the disclosure centers on two races, the practical impact extends beyond the affected contests. Polling results are often reused by newsrooms and campaigns as benchmarks for campaign strategy, message testing, and resource allocation. When polling is later found to be invalid or fabricated, it can require downstream corrections across reporting outlets that treated the surveys as credible market indicates.

The report did not indicate in its description the full scope of the problem beyond the two identified polls, nor did it lay out any specific regulatory or criminal consequences in the available account. However, the episode may raise questions about what internal controls existed, what documentation accompanied the releases, and what steps, if any, the firm plans to take to prevent future releases of fabricated results.

Election integrity advocates and election administrators generally rely on a range of data streams, including official voting records after Election Day, but polling occupies a different role by shaping perceptions and expectations before votes are cast. When polling is found to be fabricated, it creates a credibility gap not only for the firm involved, but for the broader ecosystem that consumes polling outputs.

For media organizations and other users of polling data, the episode highlights the importance of verification and transparency practices, including how polling results are sourced, labeled, and archived. It also underscores the potential need for clearer documentation of a firm’s methodology and access to underlying information when serious questions arise.

As the matter develops, the key questions will be what exactly Median Strategies publicly released, what the company used as the basis for those releases, and whether any further polling products beyond the two races are implicated. The Hill’s account frames the disclosure as a new trust challenge for an industry already facing scrutiny, suggesting that additional details may come from the company directly, as well as from outside review or reporting.