People searching on Facebook for footage of Saturday’s racist shooting in Buffalo, NY, may come across posts with footage of the attack or links to websites promising full video of the gunman. Interconnected between those posts, they may have seen different types of ads.
The social network sometimes has ads next to posts offering clips of the video, which a gunman streamed live on the video platform Twitch as he killed 10 people. Over the past six days, those livestream recordings have been circulating across the Internet, including Facebook, Twitter and Fringe, and extremist message boards and sites, despite attempts by some companies to remove the content.
The rapid pace at which the 18-year-old gunman’s fleeting livestream has expanded into permanent recordings shows that major tech platforms face challenges in policing their sites for violent content.
Facebook and its parent company, Meta, rely on a combination of artificial intelligence, user reports and human mediators to track and remove shooting videos like Buffalo One. But in some search results, Facebook comes across links to websites hosting clips next to violent videos or ads.
It is not clear how often ads appear next to the post with the video. The search for words associated with the shooting footage includes tests conducted by The New York Times and Tech Transparency Project, an industry watchdog group, with advertisements for horror film, clothing companies and video streaming services. In some cases, Facebook recommended certain search terms about Buffalo Gunman videos, noting that they are “now popular” on the platform.
In a search, the platform saw an ad for a two-post video game company, described below as a shooting clip uploaded to Facebook, described as “very graphic. Buffalo shooter.” The Times does not disclose specific words or phrases used to search on Facebook.
Augustin Foe, a cyber security and ad fraud researcher, said that large tech platforms have the potential to demonetise the search around tragic events. “It’s technically simple,” he said. “If you choose to do that, one person can easily demonetize these conditions.”
“Our goal is to protect those who use our services from this horrible content, even if the bad guys are dead to draw attention to it,” Andy Stone, a meta-seer, said in a statement. He didn’t pay attention to the Facebook ads.
Facebook also has the ability to monitor search on its platform. Searching for words like “ISIS” and “genocide” leads to graphic content warnings that users must click before seeing results.
While searching for similar words about Buffalo Video on Google did not result in any ads, Mr. Foi said there is an obvious difference between the search platform and Facebook. On Google, advertisers can choose which keywords they want their ads to show against, he said. Facebook, on the other hand, places ads in a user’s news feed or search results that it believes are relevant to the user based on Facebook’s interests and web activity.
The Google team, led by Michael Eckmann, stated that the company had designated the Buffalo shooting as a “sensitive event,” meaning that no ads could be placed against its related search. “We do not allow ads to run against relevant keywords,” he said.
Facebook has come under fire in the past for ads appearing next to right-wing extremist content. Following Jan. 6, 2021, Riots at the US Capitol, BuzzFeed News found out that the platform was posting ads for military gear and gun accessories next to a post about the coup.
Following that report, the company temporarily suspended advertisements for gun accessories and military gear through a presidential inauguration that month.