How a Troll Farm Manufactures Outrage in Eleven Days
November 15, 2026, COMMENTARY–I’m sitting here with my coffee, and I’ve been going down a rabbit hole for the last several hours that I want to walk you through, because it matters for every single one of us trying to separate fact from fiction on this platform right now.
It started with a post. DIN Intel Network. An account that didn’t exist eleven days ago. Blue check. 5,000 followers. Zero following. One follow. And on November 14 at 5:05 in the morning Pacific, it dropped a six minute video of Senator Ruben Gallego and Congresswoman Adelita Grijalva confronting House Speaker Mike Johnson outside his Capitol office during the government shutdown. Gallego telling Johnson to stop covering up for pedophiles. Johnson calling it a publicity stunt. Over 700,000 views. 60,000 likes on a clip that size from an account that size.
My first instinct was: this is important. This is real. This is breaking.
Then I started pulling the thread.
The video is from Wednesday, November 11. That’s what DIN claimed. But a reply from an account called MarketPulse pointed out something that should have been the headline: this confrontation happened in October 2025. Thirteen months ago. DIN took old footage, slapped a fresh timestamp on it, and sold it as breaking news on the eighth day of a shutdown that hadn’t even started when the video was filmed.
One honest reply in that thread. Buried under bot-grade engagement.
The Clean Room Build
So I asked the next question. Whose account is this?
Launched November 3. Eleven days old. The bio is generic. The location field just says the word Din. Not a city. Not a country. The banner and avatar were uploaded the same day it went live. No posting history. No gradual growth. No network of mutual follows building over months like a real account does. It’s a clean room build.
And every single post ends with the same line: Follow me up at DIN Intel Network, the next drop will be shocking. This one-line, crafty hook at the end of the post is presented as a signature. It’s a sales line. Repeated like a mantra across dozens of posts.
Then the numbers stopped making sense. On November 5th and 6th, the account’s first real posts pulled zero to 2 likes and a few hundred views. Then on November 7th onward, everything jumps to 8,000, 10,000, and 26,000 likes overnight. The Sarah McBride post hit 60,000 likes when the account had about 5,400 followers. That’s a 12 to 1 ratio. Impossible organically at that scale. You either buy engagement or you activate a pre-seeded follower base in waves. Either way, somebody’s paying for it.
That ratio is the fingerprint of a psyop. Not the cartoon version with fake accounts and obvious bots. The classic playbook. Seed a disposable account, flood it with engagement to manufacture the appearance of organic interest, then let the algorithm do the distribution. The account doesn’t need to be believed. It just needs to be seen by enough people that the false premise enters the bloodstream before anyone checks the date on the footage. And that’s exactly why it didn’t pass the sniff test. Real journalists don’t post six-minute clips with a thirteen-month-old timestamp and expect nobody to notice. The speed of the engagement wave was the tell. Attention becomes evidence, evidence becomes reach, reach becomes the story. When someone like MarketPulse flags the thirteen-month-old footage, that reply is the one thing that breaks the loop, which is exactly why it gets buried.
Content-wise, it’s a shotgun. Anti-Trump. Anti-Democrat. Anti-Israel. Pro-Epstein files. All in the same feed. No consistent lane. No editorial voice. No through line except outrage. This make-shift intel network is content mill under its cover, with the goal of testing which outrage converts.
The Anchor Account and the Cluster
But here’s where it got interesting. I cross-referenced with an independent amplification tracking site called Verificator, which ran a network analysis on November 8th for the hashtag Iran. They found DIN and another account called Deep Leaks HQ together accounting for 9.8 percent of all organic retweet connections in that sample. And those two accounts sit clustered right next to each other in the graph.
Deep Leaks HQ was created on May 5, 2026, six months before DIN. It has 208,000 followers but is following only eight. Same bio pattern: decoding patterns, truth over narrative, beyond headlines. Same blue check. Same zero mutuals footprint. Same clean room architecture, just with a head start and a harvested audience.
The content overlap is near total. Deep Leaks posted the Epstein files must be released angle, the anti-Israel lane, the ICE abuse clips, and the Trump impeachment material. It’s the exact same inventory DIN is now running, just with bigger numbers because it has the audience.
On November 6th, Deep Leaks posted the following promo. “Capitol Fox is back. The silence is over. The signal is live again. For those who question the narrative, dig deeper. Let’s follow and reshare his new account.” That’s not commentary. That’s seeding. Actively recruiting another node into the network before the first one gets burned.
On November 12th, Deep Leaks hit its 200,000 follower milestone with a WWG1WGA hashtag post. That’s not a community milestone. That’s a recruitment checkpoint.
And the monetization funnel is sitting right there in the bio. A Telegram link. t dot me slash unveil 444. Deep Leaks posts crypto reset panic content, like “Trump secretly replaced the Federal Reserve…gold prices…Buy Bitcoin before the system resets.” Then follows with this post, “I’ve been flooded with DMs. Send Guide to unveil 444 on Telegram for easier access.” That’s not journalism. That’s lead generation dressed up as intel. They’re farming a following on high-heat political outrage, then converting the audience into Telegram subscribers who get funneled into crypto scams or paid narrative packages.
Deep Leaks’ September 18th post claiming Netanyahu raped Virginia Giuffre pulled 81,000 likes and 1.5 million views. That’s the kind of high-heat content that builds the following DIN is now trying to replicate from scratch. The Netanyahu claim was never substantiated. It was engagement bait. And it worked.
So the architecture looks like this. Deep Leaks is the anchor account with the harvested following and the monetization funnel. DIN is the disposable test vehicle. Launched cold. Blue checked immediately. Posting recycled footage with fake timestamps to see which outrage lanes convert fastest before the real narrative gets deployed. Capitol Fox may be a third node, seeded on November 6.
The other accounts that showed up in that Verificator Iran sample. Little Think Tank, America First Graphs, BeamTracker. Those are legitimate niche accounts. French political commentary. US nationalist infographics. Military aviation OSINT. Different story. They just happened to be active in the same topic window. The real cluster is DIN, Deep Leaks, and possibly Capitol Fox.
The Older Machine: Selective Editing and Manufactured Crisis
This is not a new tactic. DIN and Deep Leaks are just the latest disposable nodes in a much older machine. Established accounts and networks aligned with Democrat messaging have been running the same playbook for years: take old footage or selectively edit clips to manufacture a crisis, vilify conservatives, and farm engagement for political gain.
The Covington Catholic video in January 2019 is the textbook case. A short, selectively edited clip of high school boys in MAGA hats standing near a Native American activist was amplified within hours by major progressive accounts, journalists, and media outlets. The narrative was locked in. The boys were aggressors, racists, and the face of Trump’s America. The full video, released later that day, showed the opposite sequence of events. Lawsuits followed. Settlements were paid. The original framing still circulates years later because the correction never travels as far as the manufactured version.
Occupy Democrats and similar long-running progressive accounts have operated this way for more than a decade. They specialize in clipping conservative speeches at the exact moment that sounds worst when stripped of context, adding a caption that invents intent, and pushing it into the viral layer. The same pattern appears with selective editing of Trump rally footage, DeSantis press conferences, and state-level Republican events. The goal is never accuracy. The goal is the 48-hour window where the outrage is useful before the full context arrives.
Media Matters and the broader progressive monitoring ecosystem have formalized the process. Researchers watch hours of conservative content, extract the most usable 15 to 30 seconds, package it with a framing document, and distribute it to aligned journalists and large accounts. The result is coordinated amplification that looks organic because the original clip is real. The deception is in what was left out and how the remaining seconds were labeled.
This is the manufactured crisis machine. An event is identified or staged. A selective edit is produced. The edit is seeded through established accounts that have spent years building audience trust. The audience reacts. The reaction becomes the story. By the time the full footage or the missing context appears, the political utility has already been extracted. The farming is not a side effect. It is the business model.
What This Means
Here’s what this means for you. If DIN is a narrative testing platform, the Gallego clip was never about Gallego. It was a probe. Drop a high emotion Epstein adjacent video. Watch which lanes light up. Which accounts quote tweet. Which hashtags absorb it. The MarketPulse reply flagging the thirteen-month-old footage is the one honest signal in that entire thread, and it got buried under manufactured engagement. That’s the tell. When the truth gets downvoted and the lie gets amplified, you’re not looking at an audience. You’re looking at an operation.
This is the part that should concern every person on this platform who still believes in finding truth through the noise. These accounts don’t need to be believed. They just need to be seen. 700,000 views on a fake timestamp is 700,000 people who walked away thinking they witnessed something that never happened that day. Multiply that by every account in the cluster, every recycled clip, every manufactured wave, and you start to see why the information environment feels so poisoned. It’s not an accident. It’s architecture.
I’m not saying every account with a blue check and a big following is a troll farm. I’m saying the pattern is checkable. Launch date. Following count. Engagement ratio. Footer text. Telegram funnel. Content scatter. Timestamp accuracy. Selective editing. Run those checks on any account that suddenly appears with massive reach, or any long-running account that repeatedly pushes the same style of clipped outrage, and you’ll know within minutes whether you’re looking at a journalist or a node in the machine.
Coffee’s getting cold. But the work isn’t.
James Bradley | Uncensored Beat
Investigative journalism. No handlers. No script.