Skip to main content

NYU researchers say Facebook advertisers lied about their identity

NYU researchers say Facebook advertisers lied about their identity


Many ads sidestepped the platform’s disclosure requirements

Share this story

illustration by Alex Castro / The Verge

A yearlong analysis of Facebook’s Ad Library has revealed “significant systemic flaws” in the way the platform monitors and enforces its political ad rules, according to researchers at New York University.

The issues were uncovered as part of the NYU team’s audit of the Ad Library between May 2018 and June 2019, which found some $37 million worth of ads did not accurately disclose who was paying for them.

Researchers Laura Edelson, Tobias Lauringer, and Damon McCoy found more than 86,000 Facebook Pages that ran political ads with the misleading disclosures, the report states. More than 19,000 ads appeared to be paid for by “likely inauthentic communities,” with groups of Pages promoting near-identical images or messages targeted at swing state voters.

The inauthentic ads used “disinformation tactics similar to those employed by the Russian-backed Internet Research Agency,” including targeting readers by race, gender, union membership or veteran status,” the report states. The Internet Research Agency is believed to be behind efforts in 2016 to hack emails from the Democratic National Committee.

A Facebook spokesperson told The Verge in an email that “our authorization and transparency measures have meaningfully changed since the research was conducted,” adding that the social media platform offers “more transparency into political and issue advertising than TV, radio or any other digital ad platform.”

Three identical Facebook ads from different Pages
Three identical Facebook ads from different Pages
NYU Tandon School of Engineering

The vulnerabilities in Facebook’s ad system could have allowed “a malicious advertiser to avoid accurate disclosure of their political ads,” the researchers found.

Facebook relaunched its Ad Library last March, saying it was “committed to creating a new standard of transparency and authenticity for advertising.” The update to the online repository, first introduced in May 2018, made all Facebook’s active ads publicly available for review for seven years, including information about who paid for the ad and how much. Researchers and journalists were given access that allowed them to analyze information about ads in the library.

But that system was criticized for being difficult to navigate, making it an unreliable tool for tracking political advertising, the New York Times reported in July.

For its part, Facebook says it’s reviewing the Pages mentioned in the NYU researchers’ report. When a Page hides its ownership to mislead people, Facebook requires additional information and a verification process for the page to remain active. Fewer than 10 percent of ads its current Ad Library ran without the required disclaimers, but all were eventually labeled, the company says.

Facebook has tightened its requirements for political ads since the research was conducted as well; any political or social issue-ads require a Federal Election Commission or tax ID number as well as other identifying information. And additional updates to the Ad Library, including allowing people to see fewer political ads, are slated to roll out later this year.

The NYU researchers recommended Facebook “take on a more active role” in improving its Ad Library security, but praised the platform for taking steps toward better transparency, calling it “the only [ad library] that provided enough data to meaningfully study. Google doesn’t include issue advertising, and Twitter’s transparency center lists only a few hundred advertisers, compared to 126,000 pages with transparent political ads provided by Facebook.”