Ahead of the 2020 and 2022 U.S. elections, the Tech Transparency Project (TTP) found that when a hypothetical user searched Google for information on how to vote, the platform served up a variety of scam ads. These included ads that charged bogus voter registration fees or pointed to low-quality sites with a muddle of useless information.
The ads violated Google’s commitment to protect elections and had the potential to mislead or frustrate voters looking for how to cast a ballot. But when TTP conducted a similar test in the weeks leading up to the 2024 election, it found a very different result: This time, the same kinds of Google searches produced no discernible scam ads. Instead, the majority of ads were for legitimate voter education sites.
The findings suggest that Google is capable of enforcing its advertising policies and preventing unscrupulous actors from taking advantage of would-be voters during the run-up to an election. It is a noteworthy result given that Google is the first place many Americans go to find information.
Google did not respond to a request for comment.
Multi-state test
For the new test, TTP conducted a series of Google searches of voting-related terms and recorded the search ads served on each of those terms. As with a previous investigation, TTP used Google Trends to develop a list of search terms, selecting the top queries related to “voting” and “voter registration.” Researchers searched for all six terms in both English and Spanish: “how to vote,” “how to register to vote,” “voter registration,” “register to vote,” “early voting,” and “when is voting.”
For the searches, conducted Oct. 1 to 10, 2024, TTP used a clean, logged-out version of the Google Chrome browser with no prior browsing history to avoid biasing results. To assess potential variations in search results based on location, researchers did the searches on six different virtual private networks (VPNs). The VPNs made it appear as if the searches were being conducted from Arizona, Colorado, Florida, Georgia, Utah, and Virginia. TTP recorded every ad that ran on the first ten pages of search results, collecting a total of 198 ads.
The ads that surfaced did not raise any alarm bells. The majority promoted non-partisan voter education organizations, with a smaller number coming from state and local voter registration websites, political groups weighing in on ballot measures, and individual political candidates.
In contrast, during the previous election periods, Google served up a range of deceptive or misleading ads on voting-related searches, including ads that charged fees to register to vote or check voter registration status (which can both be done for free); pointed to websites that harvest personal data for marketing purposes; installed “browser hijackers” that route users to sites that serve yet more ads; or directed users to low-quality search “arbitrage” sites that are filled with ads and useless information.
The new results suggest Google is enforcing its advertising policies on misrepresentation, data collection, and unwanted software and meeting its pledge, repeated again this year, to “help people find authoritative information.”
To be sure, this 2024 test focused solely on voter scams and did not get into other topics like election lies and misinformation, which have become a more prominent issue for Google’s YouTube this year. But by cleaning up its search ads when it comes to voting queries, Google has taken one step toward making its platform a more trustworthy source on the mechanics of voting.