Skip to main content

Online researchers scramble to identify Capitol raid participants

Online researchers scramble to identify Capitol raid participants


A number of groups have begun archiving footage from the attack

Share this story

If you buy something from a Verge link, Vox Media may earn a commission. See our ethics statement.

Pro-Trump protesters trying to enter Capitol building.
Photo by Lev Radin / Pacific Press / LightRocket via Getty Images

For years, police have been warning about the dangers of crowdsourced detective work — but on Thursday morning, the Metropolitan Police Department (MPD) of Washington, DC decided to do just that. In the wake of an unprecedented mob attack on the Capitol building, the MPD has released an open call for help identifying subjects. The department has circulated a digital booklet featuring 26 pages of faces from the attack, with a hotline number for anyone who can help identify them. The FBI has made a similar request, although neither agency has produced any arrests as a result of the photos. 

The alarming nature of the attack, combined with the failure of Capitol Police to detain or process the intruders, has led to a nationwide digital manhunt. The event was captured in hundreds of photos, videos, and live streams, revealing the faces of dozens of perpetrators. Now, researchers are rushing to identify and prosecute those faces as a way to reassert control, drawing a wealth of tweets, live streams, and Instagram posts into an open-source almanac of everyone who can be reliably tied to the attack. It’s a more reliable version of the internet detective work that crashed and burned in earlier online communities — but this time, the footage is spread across hundreds of different sources, and the police are in on the job.

“We aim to collect enough information to understand what happened.”

There’s so much available footage of the rally that simply collecting it has proven to be a challenge. Since the attack began, the open-source intelligence outlet Bellingcat has been collecting first-hand media from the attack, a process they opened up to volunteers as a Google sheet when the severe scale of the attack became clear. There are now more than 100 videos and a dozen full-length live streams in the spreadsheet, all ruthlessly monitored for duplicates and any identifiable faces.

“We aim to collect enough information to understand what happened,” says Bellingcat researcher Nick Waters, who is running the project. “A major part of that will be to examine who led this attack against the Capitol. We’ve already seen multiple notable members of the far-right who were part of it.”

A parallel project has formed on Reddit’s Data Hoarders forum, where a volunteer effort has begun scraping and archiving content at a massive scale. Launched last night, a thread shows 100 tweets and more than 12GB of data added to a collectively managed torrent. 

“We are basically doing their job.”

When one commenter offered a link to the FBI’s tips page, another shot back that there was no point. “Dude, the FBI has been all over this thread for hours,” he wrote. “We are basically doing their job.”

It’s a new look for Reddit, which once represented the irresponsible side of online detective work. In 2013, the platform was at the center of a chaotic search for the perpetrator of the Boston Marathon bombing. Poring through publicly available footage of the moments leading up to the explosion, volunteer investigators ended up identifying two innocent men as perpetrators of the crime.

This time, the work seems to be proceeding more carefully and successfully. In fact, many of the people in the MPD’s digital lineup have already been identified through this kind of open-source detective work. One of the most prominent figures in the Senate chamber — seen both roaming the halls and posing on the Senate dais — was identified early by journalist Will Sommer, who recognized him as QAnon promoter Jake Angeli. Others like Jason Tankersley (a member of the Maryland Skinheads) were identified by anti-fascist activists who recognized the man from previous white supremacist rallies.

In each case, the identification was important not just as a potential lead for police but for establishing a clear record of the attack. Already, right-wing pundits and sitting members of Congress have suggested the attack was an antifa false flag, disregarding the widespread Trump paraphernalia and clear connection to the Trump rally earlier that day. But the identifications so far show how hollow that claim really is, with most identified participants having clear ties to either QAnon or white supremacist groups.

Several companies have publicly cut ties with employees spotted at the raid, but they are careful to avoid naming the employee in public statements, mindful of doxxing concerns.

In other cases, the identifications have been more chaotic. On Instagram, accounts like Homegrown Terrorists have dedicated themselves entirely to soliciting tips and identifying participants— in some cases matching incorrect names to photos. In theory, Instagram has a policy against any content that “threatens to release an individual’s private phone number, residential address, or email address” — but enforcement is spotty in practice, and the account has remained up, growing to 60,000 followers.

Facebook has pledged to actively remove video taken by raid participants, seeing it as an incitement of violence. But those platform policies have meant that much of the content used to identify the crowd is already disappearing, despite the best efforts of the Reddit archivists — sometimes from Facebook and Twitter moderation and other times from mob participants catching on to the growing dragnet. Waters says he’s already seen videos disappear from prominent streamers, and “I have zero doubt others will too.”