SciTech

Big tech is changing policy for election ads

Internet advertisements have presented new challenges for civic discourse. Most notably, machine learning-based messaging optimization and micro-targeting force highly optimized and targeted political messages on viewers, often including false or misleading claims. Big Tech has long been under pressure to address the ubiquity of unchecked, misleading information, spawning policies like Twitter’s blanket worldwide ban on all political advertising and Google’s restriction of ads so they can only target general categories, such as gender, age, and approximate location.

In early September, Facebook announced that new political ads will not be accepted for the week leading up to Election Day since there may not be enough time to contest misleading or false claims. Political ads aiming to spread misinformation about the outcome of the election, such as premature declarations of victory on November 4, will also be rejected. The company also won't allow the content of existing ads to be edited, and they will add an informational label to content that questions the legitimacy of the election, which will provide facts about the integrity of the election and voting methods. In addition, posts containing misinformation about the coronavirus pandemic and voting will be removed. Facebook will limit forwarding capabilities on their Messenger service, which has already begun on WhatsApp, the company’s messaging platform that sees mostly international usage. These contrast Facebook’s long-standing support of unfettered speech and prior decisions to not ban political adverts with false or deliberately misleading content.

In a similar vein, Google informed its advertisers on September 25 that given the unprecedented amount of votes to be counted after Election Day this year, ads referencing candidates, the election, or its outcome will not be run after polls close on November 3rd. Google’s decision is in line with its broader strategy to prevent confusion around sensitive events, which has also led to Google’s temporary ban on ads with coronavirus-related terms back in March and early April in an effort to minimize confusion regarding fake goods and price-gouging. The ban now applies to explicitly election-related ads, any ads that reference federal or state elections, and ads that run based on election-related search queries (e.g. searches of candidates or officeholders). Advertisers have been told to expect the ban to last at least a week. Google will review the situation on a weekly basis, considering factors including whether all ballots have been counted and whether there are major protests breaking out across the U.S.