His camera was already rolling when a gunman pulled into a grocery store parking lot in Buffalo, New York, in a racist attack on a black community on Saturday.
CNN reports that a live stream on Twitch, recorded from the suspect’s point of view, showed shoppers in the parking lot as the alleged gunman arrived and then followed him inside as he began a killing spree that killed 10 and injured three became. Twitch, popular for gaming livestreams, removed the video and banned the user “less than two minutes after the violence began,” according to Samantha Faught, the company’s communications director for the Americas. Only 22 people saw the attack online in real time, The Washington Post reports.
But millions watched the livestream footage afterwards. Copies and links to the newly posted video circulated online after the attack, spreading to major platforms like Twitter and Facebook, as well as lesser-known sites like Streamable, where the video was reported to have been viewed more than 3 million times The New York Times.
This is not the first time perpetrators of mass shootings have broadcast their violence live on the internet and the footage has subsequently been distributed. In 2019, a gunman attacked mosques in Christchurch, New Zealand and live-streamed his killings on Facebook. The platform said it removed 1.5 million videos of the attack in the ensuing 24 hours. Three years later, after Buffalo footage was re-uploaded and shared days after the deadly attack, platforms continue to struggle to stem the tide of violent, racist, and anti-Semitic content created from the original.
Moderating live streams is particularly difficult because things happen in real time, says Rasty Turek, CEO of Pex, a company that develops content identification tools. Turek who spoke to The edge after the Christchurch shootings, says that if Twitch were actually able to pause the stream and turn it off within two minutes of it starting, that response would be “ridiculously quick.”
“Not only is this not an industry standard, this is an achievement that was unprecedented compared to many other platforms like Facebook,” says Turek. According to Faught, Twitch removed the stream mid-broadcast but didn’t respond to questions about how long the alleged shooter broadcast before the violence began or how Twitch was initially made aware of the stream.
With live streaming having become so widespread in recent years, Turek concedes that reducing moderation response time to zero is impossible — and perhaps not the right framework in which to ponder the issue. More important is how platforms deal with copying and re-uploading the malicious content.
“The challenge isn’t how many people are watching the live stream,” he says. “The challenge is what happens to this video afterwards.” During the live stream recording, it spread like an infection: acc That New York TimesFacebook posts linking to the streamable clip generated more than 43,000 interactions as the posts lingered for more than nine hours.
Large technology companies have developed a content recognition system for such situations. Founded in 2017 by Facebook, Microsoft, Twitter and YouTube, the Global Internet Forum to Counter Terrorism (GIFCT) was established with the aim of preventing terrorist content from spreading online. After the Christchurch attacks, the coalition announced it would be tracking far-right content and groups online, having previously focused primarily on Islamic extremism. Material related to the Buffalo shooting — such as hashes of the video and the manifesto the shooter allegedly posted online — has been added to the GIFCT database, theoretically allowing platforms to automatically capture and remove reposted content.
But even if GIFCT acts as a key response in times of crisis, implementation remains an issue, says Turek. While coordinated efforts are admirable, not every company participates in the effort and its practices are not always clearly articulated.
“There are a lot of these smaller companies that essentially don’t have the resources [for content moderation] it doesn’t matter,” says Turek. “They don’t have to.”
Twitch says it intercepted the stream fairly early – the Christchurch shooter managed to broadcast on Facebook for 17 minutes – and says it’s monitoring for restreams. But Streamable’s slow response means that by the time the newly posted video was removed, millions had viewed the clip and a link to it had been shared hundreds of times across Facebook and Twitter The New York Times. Hopin, the company that owns Streamable, didn’t respond The edge‘s request for comment.
Although the Streamable link has been removed, parts and screenshots of the recording are easily accessible on other platforms such as Facebook, TikTok and Twitter where they have been re-uploaded. These major platforms then had to scramble to remove and suppress the reshared versions of the video.
Content filmed by the Buffalo shooter has been removed from YouTube, says Jack Malon, a company spokesman. Malon says the platform also “highlights videos from authoritative sources in search and recommendations.” Search results on the platform return news segments and official press conferences, making it harder to find re-uploads that slip through.
Twitter “is removing videos and media related to the incident,” said a company spokesman, who declined to be named for security reasons. TikTok did not respond to multiple requests for comment. But days after the recording, parts of the video remain, which users have re-uploaded to Twitter and TikTok.
Meta spokesperson Erica Sackin says multiple versions of the suspect’s video and screed will be added to a database to help Facebook identify and remove content. Links to external platforms hosting the content are permanently blocked.
But clips that seemed to come from the live stream continued to circulate throughout the week. On Monday afternoon, The edge viewed a Facebook post with two clips from the alleged livestream, one showing the attacker pulling into the parking lot and talking to himself, and another showing a person pointing a gun at someone in a store while she screamed in fear. The shooter murmurs an apology before moving on, and a caption superimposed on the clip suggests the victim was spared because he was white. Sackin confirmed that the content violated Facebook’s policies, and the post was removed shortly thereafter The edge asked about it.
The original clip has been cut and spliced, remixed, partially censored and otherwise edited as it makes its way across the web, and its wide reach means it will likely never go away.
Acknowledging this reality and figuring out how to move forward will be critical, says Maria Y. Rodriguez, an assistant professor at the University of Buffalo School of Social Work. Rodriguez, who studies social media and its impact on communities of color, says moderation and upholding freedom of expression online requires discipline, not just in relation to Buffalo content, but also in the day-to-day decisions platforms make.
“Platforms need some support in terms of regulation, which some parameters can provide,” says Rodriguez. Standards for how platforms identify violent content and what moderation tools they use to uncover harmful material are needed, she says.
Certain practices on the part of platforms could minimize the harm to the public, such as sensitive content filters that give users the option to view potentially disruptive material or simply scroll by, Rodriguez says. But hate crimes are not new, and similar attacks are likely to happen again. Moderation, if done effectively, could limit the spread of violent material — but what to do with the abuser has kept Rodriguez up at night.
“What do we do with him and other people like him?” she says. “What do we do with content creators?”