The Announcement Was The Mechanism

When ambiguity becomes governance

Photo by Markus Winkler on Unsplash

By the time the FBI announced it was investigating encrypted Signal chats used by immigration activists, the fear was not new. It had been living in people’s bodies for months.

The Signal notifications never stopped. ICE operations ran day and night. There was more need than there were activists, and there was a dire pull to observe, dispatch, train, look up license plates, and pack and deliver food and supplies. If we stopped, what would become of these families and their children?

The environment was high-risk and high-stakes. It didn’t help that we were always being infiltrated. Signal groups disappeared, and if you didn’t happen to be in the chat when the infiltration happened, you’d miss how to become connected with the newest group. Paranoia led people to try to figure out Signal code names, so those changed often, too.

In the streets, a constant barrage of whistles, screams, and horns drowned out our other senses. The sounds of fear, anger, and pain. Good people coordinated in real time while their nervous systems ran the math on every unknown variable, and there was no real escape.

ICE wasn’t only in the streets; it was outside our homes. Following our cars. Present at the places where we took our children. And they had already shot and killed people like us more than once. Renée Good’s blood saturated the snow for months after she’d been murdered in her vehicle by an angry ICE officer. There was no forgetting the risk of this work.

The announcement did not create the chill. It found one already there and gave it a federal address.

The line

“Megan, get off the line, now!” he barked at me angrily.

I had said my full name and showed my face on the video call. It’s not a name people forget, and he’d known me for years across advocacy circles and weekly trivia nights. It didn’t matter. The line needed to be secured, and for whatever reason, I was an unknown variable. A threat. That was all there was room for in that moment. I disconnected from the call, tears welling.

I understood it. The stakes were real. Infiltration could cost a life. His sharpness was necessary in a way I could reason through even as it landed. But I’m a person, and I wasn’t okay. The yelling and distrust found every open wound and ripped them wider.

Friends texted. Three advocates vouched for me within minutes.

Why did you do that to Megan?

I was verified and welcomed back. He’d apologized. I accepted the apology, not because it didn’t matter, but because I understood what fear does to people trying to protect something. Still, I wasn’t the same after. I had known the work was physically unsafe, but I now knew the space wasn’t emotionally safe surrounded my comrades. I could see what this pressure was doing to all of us.

If a person who’d known me for years could see me as an infiltrator and threat, I realized nobody doing this work could consistently separate the fear from the person, the moment from the pattern, and the wound from the intent. The pressure landed on all of us, with our different histories, capacities, and breaking points.

We showed up imperfectly, nervous systems constantly activated, and psychological harm multiplied. I believe the harm calcified into something that reorganized how many activists saw the people around them, and even each other.

Under enough pressure for enough time, threat becomes ubiquitous. It’s the result of nervous systems doing exactly what nervous systems do when the threat never stops.

When the threat follows you home

ICE followed us and staked out our homes as if we were criminals. When the threat becomes that personal and inescapable, everyone becomes a potential vector. The person who handled something badly, was it a mistake, or was it something worse? The person who wasn’t where they said they’d be, was it chaos, or was it deliberate?

We stopped treating human error as human error.

Apologies became admissions. Mistakes became evidence of intent. The tools of accountability, which exist for real reasons, began doing different work. What would once have been repair is now verdict. We applied intent to behavior that was probably just people failing under impossible conditions, the way people do.

I understand why it happened. I lived it. The danger was real. The need to find the source of the threat was real. But the threat wasn’t each other.

What hypervigilance costs and what it saves

People were being killed, and everyone organizing knew it.

On January 7, ICE agent Jonathan Ross shot and killed Renée Good, a 37-year-old mother of three and U.S. citizen, in her vehicle in Minneapolis. Federal officials said the agent acted in self-defense; a bystander video reviewed by Reuters, the BBC, the Wall Street Journal, and the Associated Press challenged that account.

Seventeen days later, Alex Pretti, a 37-year-old VA intensive care nurse and U.S. citizen, was shot by CBP officers after stepping between an agent and a woman who had been pushed to the ground. He was filming on his phone. He had a licensed firearm; the bystander video appears to show an agent removing it from his holster before he was shot. He died on the street. Federal authorities later labeled him a domestic terrorist. The state of Minnesota and Hennepin County subsequently sued the federal government for withholding evidence in both killings.

People were being dispatched. Immigrants were being warned of imminent ICE activity. Mutual aid networks were running real-time responses to real-time operations. The monitoring was saving lives, and everyone inside the ecosystem knew it. Every notification might be the one that mattered.

Hypervigilance produces its own reward loop. The relief of catching something in time, of dispatching help before it was too late, feels like purpose. Like the only appropriate response to a genuine emergency. Which it sometimes was.

But sustained at that frequency, over months, without rest, the same vigilance that saved people eventually couldn’t distinguish between an ICE agent and a friend. The pattern recognition that made someone a good organizer in a crisis could make them a poor judge of trust in a fracturing community. The nervous system, running a constant threat assessment, does not easily downshift into nuance.

You do not need to discredit the movement. You do not need to arrest its leaders. You need only maintain enough ambient threat, enough operational uncertainty, enough visibility of force, that the people inside the movement begin doing the work of fracture themselves.

I turned off notifications at one point to give my nervous system some downtime. I missed an ASAP response and found myself outside the Signal groups entirely, which shifted names and membership constantly for security reasons, I understood. I didn’t chase verification. I kept helping, raising money, writing cards to families, distributing food, doing the work that actually reached the people the movement existed for, outside the Signal chat infrastructure entirely.

When I turned off Signal notifications, I felt like I wasn’t doing enough. The cost of self-preservation was guilt. Staying connected meant burnout. Stepping back meant guilt. There was no clean exit.

The announcement itself was the mechanism

Kash Patel announced the FBI was investigating encrypted Signal chats used by immigration activists. No specific crime was named. Signal is end-to-end encrypted, and investigators cannot easily access its contents. Its effect was atmospheric, not legal. And the disclosure was psychological, not investigative.

By framing the inquiry around whether sharing information about ICE movements could constitute interference, federal authorities signaled that coordination itself might now be viewed as suspicious behavior. The announcement found people already afraid of each other and handed that fear a federal address. Uncertainty did the work that charges could not.

In the days that followed, Signal chats shifted names. People were booted or re-verified with more fervor. The architecture of organizing, already strained by months of crisis, bent further under the new weight. People who had once trusted each other with their locations, families, and safety plans now ran threat assessments on the same people.

The federal government didn’t cause that, but it did accelerate something already in motion.

The same logic had appeared in New York before it landed in Minnesota. In late 2025, reporting revealed the FBI's monitoring of immigration-rights CourtWatch Signal chats, where legal observers documented immigration proceedings and law enforcement behavior inside courtrooms.

Federal documents reportedly characterized some participants as anarchist violent extremist actors. Their activity: watching and taking notes. Recording what happened in public proceedings. Simple observation had become its own evidentiary category.

Journalism becomes participation

The same logic that reframed coordination as conspiracy reframed documentation as involvement.

Federal authorities charged Don Lemon, the former CNN anchor, and eight co-defendants in connection with a January 18 protest at Cities Church in St. Paul, where demonstrators alleged that one of the church’s pastors served as an ICE field director. Lemon and independent journalist, Georgia Fort, maintained they were present solely to document the protest.

The charges, conspiracy against religious freedom and interfering with the exercise of religious freedom, had never previously been applied in a protest context, according to a Justice Department official’s own concession in court. A federal magistrate judge had earlier rejected the request for an arrest warrant, finding no evidence of criminal behavior. A grand jury later indicted. Lemon pleaded not guilty in February 2026.

The framing implied something specific: that documentation could be reinterpreted as involvement, that presence could be retroactively read as disruption, that intent could be inferred after the fact.

The 2026 Counterterrorism Strategy makes the same interpretive move at the federal level. It formally names anti-fascists alongside cartels and jihadists as national security concerns. But anti-fascist is not a membership category. There is no organization to join or leave. It is a political disposition, opposition to fascism, that has existed across decades and ideological traditions. Designating it a security threat does not target an organization. It targets an interpretation.

“Anti-American” functions the same way. Unlike legally precise criminal language, it is an identity-adjacent descriptor, broad enough that people begin running uncertainty calculations around whether ordinary dissent, protest, or association might someday be read through that frame. The shift matters not because violent acts should go unaddressed. Violence can be investigated as violence. The danger is when opposition to fascism itself becomes treated as a suspicious identity category. And interpretive categories expand easily under political pressure.

Once the burden of proof shifts onto the identity rather than the act, participation itself becomes evidence.

No law changed. What changed was the uncertainty about where the line was. And uncertainty, in a climate where federal operations had already turned deadly, is enough.

Journalists began exercising greater caution about proximity to events because the cost of being wrong had become incalculable.

That gap between unchanged law and changed behavior is the proof.

What is actually being tested

This moment is testing whether a democracy can be quietly taught to accept ever-narrower definitions of legitimate civic behavior. Will people stop organizing on their own, and will journalists step back without being banned? Will uncertainty replace overt coercion?

The evidence, so far, suggests it can.

And underneath all of it, the people the movement existed to protect are still there. They continue to be detained and separated from their families. There is still an imminent need for mutual aid.

Immigrants didn’t stop needing help because the activists stopped trusting each other. The crisis didn’t pause for the infighting. It continued, a little more quietly, with a little less capacity responding to it. And somehow even pointing that out in some circles is viewed as harm.

White progressive spaces can be especially vulnerable to this failure: turning the struggle inward until the emotional management of the helpers competes with the needs of the people at risk. The movement turned inward at exactly the moment it needed to hold outward.

The people most at risk kept waiting.

Naming the mechanism

This is democratic erosion. A recalibration and narrowing of civic space. Stress is applied slowly enough to feel administrative, procedural, and technical rather than political.

The underlying engine is probabilistic punishment: a mechanism that prompts people to alter their behavior once they can no longer confidently determine where punishment begins. Uncertainty becomes the instrument. Prosecution does not have to be likely. It only has to remain possible enough that people cannot rule it out.

That is what links every element of this piece: the elastic category language, the Signal investigation announcement, the charges against journalists for presence, and the monitoring of courtroom observers. Each example shows punishment operating before conviction. The threat does its work earlier, at the point where ordinary participation starts to feel legally and personally incalculable.

Civic chills are reversible precisely because they operate through uncertainty rather than law. They weaken when ambiguity is named, when boundaries are forced into the open, and when lawful civic behavior resumes visibly with institutional backing.

Reversal is possible, but it has to happen before uncertainty hardens into precedent. Otherwise, what begins as behavioral adaptation inside movements hardens once institutions adapt, too.

Megan M. American Horse writes about democratic backsliding, state power, and the quiet mechanics of intimidation that make people censor themselves before punishment ever arrives.