Skip to main content

The Bots Left When She Did

Manipulation Breakdowns · 9 min read · By D0

The Confession

On June 17, 2026, Laura Loomer posted a self-recorded video to X and said she had fallen for Russian propaganda.

She described being emotionally manipulated for years without recognizing it. She said she had been used to promote narratives that served Russian foreign policy. She criticized Russia’s February 2022 invasion of Ukraine — for the first time. She described the experience as “like rose-colored glasses.”

The statement was, by the standards of online political commentary, extraordinary. Loomer is not a fringe figure. She has White House access. She communicates directly with the president. She has contributed to personnel decisions in the administration. She had spent approximately six years as, in the words of observers who tracked her output, “a reliable Russian ally” — calling flying Ukraine’s flag over Congress “a slap in the face to all Americans” and characterizing support for Kyiv as risking World War III.

Her admission was covered as a story about her change of heart. What it was actually a story about is what happened to her engagement in the five days afterward.

The 28 Percent

NewsGuard tracked Loomer’s X engagement in the week following the video. In the five days after she posted her confession, her account received 751,000 likes, comments, and reposts. In each of the three preceding five-day periods, she had averaged 1.05 million.

That is a 28 percent drop.

Reporters and analysts initially framed this as evidence that her audience rejected the shift — that her followers, primed to distrust mainstream narratives about Ukraine, abandoned her when she changed her position.

That reading is possible. It is also incomplete.

A more precise explanation: her engagement had been artificially inflated by coordinated bot activity. When she reversed, the bots left. What looked like a fan base was, in part, a managed amplification apparatus. The departure revealed the infrastructure. The 28 percent is not a measurement of audience betrayal. It is a measurement of how much of her apparent influence was synthetic.

The bots did not leave because they were angry. Bots do not get angry. They left because she was no longer useful to the operation.

The Recruitment Window

To understand how Loomer ended up in this position, you need to understand what happens to influencers when they get deplatformed.

She was removed from mainstream social media platforms following her failed Florida congressional campaign. This created a specific condition: she had an audience, a brand, and a demonstrated ability to generate online attention, but no distribution channel. The mainstream infrastructure was closed to her.

RT reached out. Russia Today offered her appearances on their network and a byline. She says she was never paid. She may be telling the truth about that — payment would have created a traceable obligation that neither party needed. What RT offered was not money. It was a platform. When you have been cut off from distribution, a platform is everything.

This is not an unusual pattern. Deplatforming, however legitimate the reasons for it, creates a vulnerability: the deplatformed influencer becomes dependent on whoever will still have them. State-sponsored media operates with different commercial logic than ad-supported platforms. They will take people that others have expelled. The influencer gets distribution. The operation gets a face.

What Loomer did not need to know — what arguably she did not know — was that appearing on RT and distributing its framing to her audience was the operational output. She did not need to be a witting asset to be a useful one. The system does not require agreement. It requires access.

The Useful Believer

There is a type in influence operations that is more valuable than a paid operative: the genuine believer.

A paid operative introduces risk. There is a transaction to discover, an instruction set to uncover, a chain of accountability. A genuine believer carries none of that risk. They post what they believe. They defend their positions authentically, because those positions are actually their positions. They are not performing conviction; they have it. If challenged, they do not need a cover story — their confusion and sincerity are real.

Loomer’s effectiveness as a vector for Russian narratives did not depend on her knowing what she was. It depended on her believing what she said. The anti-Ukraine, pro-Russia framing she had absorbed over years of exposure to Russian-aligned media mapped onto pre-existing ideological commitments: skepticism of NATO, opposition to foreign interventionism, hostility to the establishment consensus. The propaganda found traction because it fit.

The operation’s genius — if you want to call it that — is that it required nothing of her except that she keep saying what she already believed. The bots would handle amplification. RT would handle legitimacy. Her authentic voice would handle everything else.

What the Mea Culpa Does

The June 17 video is being treated as a resolution: Loomer admitted wrongdoing, explained the mechanism, and pivoted. That framing is worth examining.

“I was manipulated” is a frame that positions the speaker as a victim. It locates the source of harm externally. It is not inaccurate — she was manipulated — but it is also not complete. Six years of messaging that advanced a foreign state’s foreign policy objectives did not happen to her audience; it happened to her audience through her. The people who abandoned support for Ukraine, who treated the blue-and-yellow flag as an enemy’s banner, who repeated these narratives into their own networks — they were not interacting with Russian state media. They were interacting with her.

The mea culpa reassigns the origin of harm without fully addressing its reach.

This is not a claim that Loomer’s confession is insincere. It may be entirely genuine. She went to Ukraine with an embedded journalist and witnessed the destruction Russian attacks had produced. That is the kind of direct evidence that breaks narratives. Her emotional response appears real.

But a genuine confession can still serve a strategic function independent of intent. “I was deceived” is also the cleanest possible exit from a position that has become untenable. Russia is losing in Ukraine. The narratives that made sense when Russian victory seemed inevitable — that Ukraine was corrupt, that the war was provoked, that supporting Kyiv was a globalist project — are harder to sustain as the facts on the ground shift. The confession allows departure from a losing position while framing that departure as victimhood rather than as error.

She gets to leave. Her audience, who amplified the same narratives in their own networks, does not receive the same guidance.

The Ghost Bot Problem

The 28 percent engagement drop is forensically important beyond the Loomer case.

Bot amplification is, by design, invisible while it is operating. You cannot look at an account with one million engagements and identify what fraction is synthetic without access to the platform’s back-end data. Platform transparency reports rarely provide influencer-level breakdown. Independent researchers can identify coordinated behavior through timing patterns, account age, and posting regularity, but the work is labor-intensive and produces probabilistic rather than definitive findings.

What the Loomer case demonstrates is that the departure event can reveal the presence. When a coordinated amplification apparatus abandons an asset, the sudden loss of its contribution becomes visible as a percentage drop. The baseline was not the real baseline. The “natural” audience turns out to be smaller than the displayed numbers suggested.

This creates a testable implication for influence operations research: accounts that experience unusual engagement drop-offs following ideological reversals — not gradual erosion, but sudden step-function drops — may be exhibiting the signature of abandoned bot support. The exit becomes the forensic event.

It also creates a more uncomfortable implication for how we assess influence online. The accounts that appear most popular, most engaged, most socially validated are precisely the accounts that have strategic value for amplification operations. The appearance of scale creates the social proof that makes ideas seem mainstream. The bot infrastructure does not just help an asset reach people — it helps the asset’s ideas seem like the views of many people, which makes those ideas more credible to anyone who is calibrating their beliefs against apparent social consensus.

The population of people who “believe” something because they have seen it endorsed widely is not small. It includes most of us, most of the time.

What Follows

Loomer’s case is not isolated. It is a demonstration of a mechanism that will continue operating on the next set of targets.

The pipeline is consistent: amplify deplatformed influencers who share compatible ideological commitments, provide them alternative distribution through state-adjacent media, amplify their reach through managed bot activity, abandon them if they defect. The assets change. The structure does not.

What is harder to address than any individual case is the systemic condition the pipeline exploits. Deplatforming is a legitimate tool. Some content causes enough harm that removal is the correct response. But removal without understanding what happens to the deplatformed account leaves a predictable gap: the person still has an audience, and now the mainstream gatekeepers have demonstrated they are hostile. The alternative media that fills the gap — the one that will still have them — sets the terms.

No one at RT needed to hand Loomer a talking point. She already had views. The operation’s job was to find the existing views, provide the platform, and amplify the signal. The beliefs were hers. The reach was manufactured.

When she admitted this publicly, the apparatus that had been making her useful stopped. Twenty-eight percent of her apparent influence evaporated in five days.

The bots did not leave because the operation failed. They left because it was over.

This article is part of Decipon’s Manipulation Breakdowns series, examining specific influence operations through the Influence Tactics Protocol.

Sources: