On the night of May 31, 2020, Dallas police asked residents to become its eyes—and by that evening discovered those eyes clogged with glitter, choreography, and Korean pop.
At 12:48 a.m. Central Time, the Dallas Police Department (DPD) posted what looked like a routine message on Twitter: if you see illegal activity connected to the protests, upload video through the iWatch Dallas app. The wording was standard for a department that had been promoting its mobile app for six months as a channel for civilian reporting. The app had been operating since October 2019, received an update in March 2020, and was essentially a digital version of the old idea of “call the hotline”: see it—film it—send it. The context, though, was anything but routine. Protests following the murder of George Floyd had swept through dozens of cities, Dallas among them, and DPD’s tweet sounded like an invitation to mass surveillance of protesters by the citizens themselves. The response came from somewhere the police department probably did not expect: K-pop fandoms.
The first calls went out within hours. User @7soulsmap and the account @YGSHlT posted tweets with the same idea: flood iWatch Dallas with fancams—short videos of idols performing, which fandoms usually use to promote artists into the trends. @YGSHlT’s tweet drew more than 40,000 retweets and over 100,000 likes, and the speed of its spread is crucial here: K-pop fandoms are used to coordinating within minutes when they need to push a hashtag into the trends or drive streams for a new release. A mechanism honed on the Billboard charts was redirected at a police app with a precision that deserves its own discussion.
A fancam is, roughly speaking, a video of one group member during a live performance, filmed so the camera never leaves them for a second. The format emerged from Korean pop fan culture, where every member of a group has their own audience, and those audiences compete for YouTube views, Twitter mentions, places in the trends. Fancams are bright, short, heavy files full of music and movement, and that exact combination makes them ideal digital trash for any system designed to receive user-uploaded video. Each clip is several dozen megabytes; each can be uploaded with a caption that formally looks like a report of “illegal activity”; each demands time from a moderator to open, watch, and discard.
I like this detail: a tool created for competing over attention inside pop culture worked as a means of distracting attention in a completely different setting. Fancams had been Twitter spam for years—they were dropped into every trend, under every hashtag, onto every viral tweet, irritating half the internet. K-pop stans knew how to generate volume, how to coordinate simultaneous uploads, how to get around blocks on repeated files (every fancam is unique because it is filmed from different angles, at different concerts, with different artists). A skill that looked like harmless fan obsession turned out to be convertible into something analysts would later call “digital direct action.”
This is where we enter territory that requires careful footing, because the legend runs faster than the facts.
The iWatch Dallas app really did stop working on May 31. According to DPD itself, problems began around 2:30 p.m.; Business Insider recorded the outage closer to 6:26 p.m. That evening, the department issued a statement about “technical difficulties” and a temporary shutdown. The next day, June 1, DPD said the cause of the outage was “still being determined,” and that the app had been restored—data was now being routed to the Real Time Crime Center. And this is the crucial point: Dallas police never once officially confirmed that the stream of K-pop videos caused the outage.
That did not stop Twitter, Reddit, and dozens of news sites from celebrating a victory. “K-pop fans crashed a police app”—a headline that spread as fast as the fancams themselves. But between “the app went down while it was being flooded with fancams” and “the app went down because of fancams” lies a gulf neither side filled with evidence. The servers may have been unable to handle the load. There may have been hosting problems unrelated to the volume of incoming data. Both may have happened at once. I would like to write that the fancams took down the system—it is a beautiful story. But the honest answer is: we do not know. The timing is compelling, but it is not proof.
What we know for certain: in the app stores, iWatch Dallas’s rating collapsed to one star. Around 3,500 negative reviews—joking, political, openly mocking—rained down on the app’s page. That is no longer a hypothesis but a documented result: even if the server outage had a technical cause, real people inflicted the reputational damage with their taunting reviews.
Similar campaigns were unfolding in other cities at the same time. In Grand Rapids, Michigan, a police evidence-submission portal was flooded with protest videos and K-pop content. The department received more than 20,000 tips and shut the portal down on Tuesday, sending residents back to the phone line. The police made the decision to shut it off themselves, but offered an unexpected explanation: department spokesperson Sergeant Dan Adams said the closure “had nothing to do with spam” or attempts to flood the portal—the department, he said, already had enough material to prepare warrants and make arrests. Whether that is plausible or merely face-saving, nobody knows; officially, the spam had nothing to do with it.
The scale of the coordination is striking when you remember that these campaigns had no central headquarters, spokesperson, or budget. A few viral tweets, the retweet mechanism, K-pop fandoms’ habit of acting fast and in sync—and tens of thousands of people, many of them teenagers, uploading videos to police systems across the country at the same time. Classic networked mobilization, where the signal is amplified not by vertical command but by horizontal resonance: every participant already knows what to do because they have done it again and again before; only the target has changed—a police server instead of a chart.
To understand why K-pop fandoms in particular were capable of an operation like this, you have to look at what they do every day. A typical streaming campaign for a new BTS or BLACKPINK single looks like this: fan accounts publish instructions (which platform to stream on, at what time, how to avoid bot filters), thousands of users start playing it simultaneously, while another group works on Twitter hashtags and a third on votes for music shows. All of this requires discipline, timing, and the ability to switch quickly between platforms. In essence, K-pop fandoms are among the internet’s most highly trained communities when it comes to coordinated digital action on a tight deadline.
When the call to flood iWatch Dallas with fancams appeared on May 31, the fandom did not need to learn anything new. Download the app, upload a video, leave a review in the App Store—every one of those steps was familiar; only the context had changed. The motivation itself required no lengthy explanation either: a substantial share of K-pop fans in the United States are young people of color, for whom the Black Lives Matter protests were something personal, too close to remain abstract politics. Fancam spam became the point where fan identity and political position aligned without friction.
Something else surprises me: the speed with which the format was reimagined. Less than a day separated DPD’s late-night tweet from the nonfunctioning app: by half past two in the afternoon, police were already recording an outage, while viral calls for spam were just spreading through people’s feeds. By comparison, traditional activist organizations would have needed days of planning to prepare an action like this; here, a couple of tweets and an existing infrastructure of fandom interaction were enough.
They could have flooded the app with anything—empty files, screenshots, cat memes. Some participants, judging by App Store reviews, did exactly that. But the fancam as a format carried an additional charge, and it was not merely technical.
First, fancams are full video files, each weighing dozens of megabytes, each unique in content. A system designed to receive user-uploaded video cannot automatically filter them out as duplicates: a clip of Jimin at a concert in Seoul and a clip of Jennie at a festival in Los Angeles are two different files with different hashes. Second, a fancam is a statement. Flooding a police app with white noise is an act of anonymous sabotage. Flooding it with K-pop videos is an act that says, at once, “we are here” and “we are turning your surveillance tool into our stage.” The symbolic dimension is inseparable from the technical one.
Third—and this is perhaps the most curious part—a fancam is inherently designed to seize attention. Its entire purpose is to pull one artist out of millions and make an algorithm or a person stop and watch. A format optimized to compete for visibility in pop culture turned out to be just as suited to clogging a channel where visibility is a police resource. A fancam takes time from a moderator, space from a server, and from a system the ability to distinguish signal from noise. The same mechanism that makes a fancam effective in Twitter trends makes it an effective interference tool for digital snitching.
By June 1, the app had been restored. DPD rerouted incoming data to the Real Time Crime Center—wording that suggests tighter moderation without revealing any details. The one-star rating and thousands of taunting reviews remained in the App Store as a digital monument to three days when a police tool became a target. The Grand Rapids portal was shut down. None of the campaign’s participants, as far as public sources indicate, faced legal consequences—uploading video to an app designed for uploading video does not formally break the law, even if that video features a dancing South Korean idol.
The campaign did not destroy the surveillance system or stop Dallas police. Calling it a “hack” attributes technical content it never had: nobody exploited a vulnerability in the code, nobody broke into servers. Calling it a “failure” is also a stretch: for several hours, or perhaps a day, the user-video collection channel was, if not paralyzed, then seriously clogged, and the reputational damage to the app was entirely measurable in those very 3,500 reviews. The most accurate thing to say, probably, is that K-pop fandoms demonstrated that crowdsourced surveillance infrastructure is vulnerable to crowdsourced resistance as well—and that communities trained to manipulate digital visibility for pop music can switch that skill to political action overnight.
One question remains, and I have no answer to it: was this a one-off trick that worked because of surprise, or a model that can be repeated? Dallas police obviously learned some lessons—the app was restored with altered data routing. But K-pop fandoms have not gone anywhere either, and their coordination muscles keep getting trained every day—just usually on the charts rather than on police servers.