How a "Predator Catcher" Crossed the Line

Dubi is known in parts of the Roblox community as a predator catcher. He runs bait accounts, compiles reports, and presents himself as someone who targets the very individuals our platform exists to identify and flag. He is also one of Rotector's top donors, which should tell you how little anyone here wanted this outcome.

He is now flagged in our system, and he has spent a little over a week appealing it. This post explains why the flag exists, what evidence we dropped along the way, and what evidence he has never been able to explain. It also documents a pattern. Every claim he made that we could check turned out to be false.

TLDR: while operating a bait account, Dubi came across a person who claimed to be 19. Before he had even stated that the bait account was 15, he immediately sent them NSFW Roblox content. This was not the usual sting. The bait account was the one constantly posting NSFW, including explicit photos of himself. When we asked why a sting needed any of that, the answers kept changing, and they have kept changing since this post first went up. Untangling them is what the rest of this post is about.

Here is the current and final flag, word for word:

The reason attached to his account. It is not about the clip, his voice, or anything we could not prove. It is about how he conducted the bait operation.

The flag changed as the case went on. During the appeal we removed every allegation that could not be sufficiently backed up. What remains is based on the conduct surrounding the sting and nothing else, supported by screenshots of his own messages.

A note before we start. We have never called Dubi a pedophile. We also withheld most of the evidence in this post from Dubi, deliberately. In a previous case, we learned that sharing evidence with a subject early in an investigation gives them material to build new stories around. We did not repeat that mistake here. He was still given every opportunity to explain his own conduct, and his responses are quoted throughout.

There is also the question of why this is public at all. For most of this, Dubi insisted he did not want it public either, telling us that going public could compromise his investigation. So we held back too, and kept the case in a closed appeal for over a week. Then he went public himself anyway, with a video about us and a threat to sue us for defamation. Rather than let his version stand on its own, we are laying out the full picture here.

A few names come up more than once, so here is who they are up front:

• Dubi: the subject, known across his accounts as officialdubi, epicdubi, and the bait account 0fv.
• Luna: a staff manager at Rotector who handled this appeal.
• Heceta: a reviewer and support staff member on the same team, who also handled this appeal.
• Acedia: an independent investigator who compiled the report at the center of the case.
• Schlep (Michael): a well-known Roblox predator catcher, here only because Dubi named him.

Everyone is referred to by the handle they use.

This post contains censored screenshots of sexually explicit conversations. We are publishing it for transparency, not as an invitation to contact or harass anyone named in it.

How This Started

Earlier this year, a short clip was sent to our team. It was about ten seconds long and carried the filename uncapped_MedalTVRoblox20240805085520. The MedalTVRoblox part and the trailing numbers match the default naming Medal gives recordings, which would put the timestamp at August 5, 2024, 8:55 AM. The uncapped_ prefix is not part of that format, so the file had already been renamed at least once before it reached us, which is worth keeping in mind.

A frame from the clip alongside its original filename. Medal recordings embed a timestamp in the file name.

The clip appears to show a round of Eviction Notice, a Roblox elimination game. A voice on the recording makes sexually inappropriate comments to a woman in the call. The in-game player list includes a player named "dubi."

The player list from the clip. A player named "dubi" appears under the jury section. We have redacted another name.

We want to be upfront about this clip, because being upfront about weak evidence is the only way the strong evidence means anything.

We cannot authenticate it. The file has been re-encoded at least once. Its original metadata is gone, its container is a QuickTime export rather than a Medal MP4, and its resolution and framerate do not match what Medal would produce on its own. We cannot treat it as an original, and its real date and origin are unknown and unverifiable. The only thing tying it to August 5, 2024 is the filename, which anyone can type or rename. The person who surfaced the clip also has credibility problems of their own, so we treated everything from them with caution.

Our team did voice comparison work between the clip and videos Dubi provided himself, including a new voice sample. We noticed a change in the audio between his recordings. When asked, he told us he had factory reset his PC, that his Blue Yeti noise suppression software was no longer active, and that the background noise was his computer fan.

We asked about the difference in audio between his new recordings and earlier clips. His answer was a factory reset and a computer fan.

The reviewers who compared the recordings found the voices similar. But voice similarity is a judgment call, and a judgment call layered on top of a file we cannot authenticate is not a foundation for anything.

So we did the only reasonable thing: we removed the clip portion of his flag for further review, and later we dropped it completely. If the clip had been the whole case, there would be no case and this post would not exist. What kept the investigation open was everything that happened after we asked him about it.

One detail is worth holding onto, though. The filename points at August 5, 2024. Later, when we confront him over the badges, Dubi writes that exact date himself. The match could be a coincidence, and we are not basing anything on it either way. Just keep it in mind for when we get there.

The Game He Said He Never Played

So why did we keep pushing on the clip? Because of what Dubi said when we asked him about it.

On May 25 he put this in writing:

"here's a google drive showing i have no dms with the female in question, that i've never played the game eviction notice which is the game in the clip, and proof that there is multiple impersonators of myself."

His May 25 statement. The claim that matters is the one in the middle, because it is the only one that can be checked.

"I never played that game" is a checkable claim on Roblox, because games award badges with visible award dates that sit on your profile for anyone to see.

As proof, he sent a screenshot of himself trying to vote on Eviction Notice and being stopped by a dialog reading "You must visit the experience before you can vote on it."

His proof that he never joined the game: Roblox telling him he must visit the experience before voting.

Luna, our staff manager, broke that proof in one afternoon. She opened Eviction Notice on her own account and got the exact same dialog. Her account has held the game's "Played" badge since January 1, 2022.

Luna got the same "you must visit" dialog on an account that has owned the game's badge for years.

Her "Played" badge, unlocked January 1, 2022. The vote dialog appears anyway.

This is a known Roblox bug. People have been posting about it publicly since at least 2023.

A forum post from November 2023 describing the same behavior: Roblox failing to register that a game was played, including the same vote error.

So his proof proved nothing. What did his profile say? The epicdubi account holds five badges from Eviction Notice: Evicted, Nominee, Safety, Leader, and Played.

The epicdubi profile. The badge row at the bottom is from Eviction Notice, the game he said he never played.

These were not sitting on the front page of his inventory. Luna found them on page 5, after scrolling through his badge pages one at a time. They are old badges, buried where you would expect badges from 2024 to be, not anything recent or displayed.

When we first told him about them on June 2, he did not explain them. He called us liars:

"out of all 13 pages on the account epicdubi there isnt a single fucking badge from eviction notice. that's just a fucking blatant lie."

Then he raised the stakes himself. To prove there were no badges, he uploaded an unlisted YouTube video titled "rotector appeal evidence #67" showing his own inventory. Luna watched it. The Eviction Notice badges were right there in his own footage.

The video he recorded to prove the badges did not exist shows the Evicted badge sitting in his inventory.

We showed him a frame from his own video next to a fresh screenshot of his profile.

Our messages on June 2. The badges were on his account and visible in the video he had just uploaded.

At that point the story changed again. He told us there was a second clip, timestamped in 2024, made by the same person, and that the whole thing was manufactured evidence.

His immediate response: "u know how i can prove this is fake rn."

He brings up a "second clip" from 2024 as proof of fabrication. We have redacted a name.

At 6:01 AM, he posted a timestamp screenshot and wrote this:

"i played this game on aug 5th 2024 which would mean i would've been like 16 saying all of this to a 19 year old woman"

A week earlier he had written "i've never played the game eviction notice." Now he wrote "i played this game on aug 5th 2024." That is the exact date embedded in the first clip's filename.

Minutes later, he deleted the screenshot he had just posted. You can see the "image is no longer available" notice in our capture below, sitting where the screenshot used to be. Our team had already saved the image itself.

The sequence on June 2. He posts an image, writes that he played the game on August 5, 2024, and the image becomes unavailable shortly after.

The screenshot he posted showed August 5, 2024, 10:54 AM. He offered to show ID, then called the clips doctored again.

That message also raises the question of his age. He claims he would have been "like 16" at the time, and he offered to show ID. We cannot verify his age and we are not going to claim a number ourselves. His Roblox accounts go back to 2008 and 2009, which would make them implausibly old for someone his claimed age, but aged accounts get bought and sold all the time, so we draw no conclusion from that either. What we can say is that the claim does not help him. It does not explain the badges.

His age claim, in his own words: by his own math, playing the game in August 2024 would have made him "like 16."

The same epicdubi account, joined 17 years ago. An account that old is hard to square with someone who says he would have been 16 in 2024, though a resold account would explain it.

He did not have to lie. He had two honest options. He could say the voice was not his, or that he was not the "dubi" in the player list. Both are hard to disprove, and an unprovable denial would have worked in his favor. Instead he denied ever playing the game, the one thing his own badges could disprove. An innocent person has no reason to lie about something that easy to check.

We dropped the clip in the end anyway, once he provided counter-evidence on the voice. But the badge lie changed how we treated everything he said after that, and it is the reason we kept investigating instead of closing the case.

The Document

With the clip dropped from the flag, we were still left with the sting conduct. This is the part of the flag that was upheld, and it is the reason this post exists.

In late May, a 12-page report reached us. It was compiled by Acedia, an independent investigator who runs a small volunteer group that tracks condo activity on Roblox and documents the users behind it. He had received a tip about the conduct and put together a report on it. The document covers a Discord account named 0fv and its interactions with another user, the account links tying 0fv back to Dubi's main accounts, and an investigative addendum arguing that the conduct on display was active participation rather than passive observation. You can read a censored version of the report here.

One thing to get ahead of, because Dubi builds a defense on it later: he eventually got Acedia to take this document down, then waved that retraction around as proof the whole thing was fake. It was not, and we walk through exactly what happened with it further down. The report itself does not change either way.

Dubi does not dispute that he operated the 0fv account. His defense is that everything it did was part of a honeypot. Most people who actually run these call them stings, but the label matters less than the rule behind all of them, which is that you document what the target does and never take part yourself. Documenting and taking part are not the same thing.

There is a word for breaking that rule, and since it comes up in this case, it is worth defining properly: entrapment. Entrapment is when the person running the operation induces someone into a wrong they were not already committing on their own, by initiating it, suggesting it, escalating it, or supplying the material for it. The formal legal defense only applies against law enforcement, but the principle applies to anyone running bait. The moment the operator is the one driving the misconduct, the operation stops producing evidence of what the target would have done and starts producing evidence of what the operator made happen. Any case built on it collapses. That is why real sting operators are obsessive about keeping their decoys passive.

The 0fv account did the opposite. The evidence shows repeated direct participation: Dubi engaging with this person and sending explicit content himself, instead of sitting back and letting the other side incriminate themselves. That is the issue that survived after every other allegation was dropped from the flag. We will call it what it was from here, a sting, and a poorly run one.

Here is what Dubi actually did with the 0fv account, in order, on February 26, 2026. Every screenshot below comes from the report, with explicit material censored. Since he does not dispute that he was the one typing, we are going to say "he" instead of "the account."

He made the first move.

The start of the DM history. 0fv messages first: "hi" and "sorry for it taking so long."

The flirting came first, then a question about age. The person on the other end is the one Dubi calls the predator. He addressed that person as "maam," said "I have a taste for older women, that's why I'm asking," and asked their age. They answered: 19. He replied "thank god." Then the explicit material started.

0fv asks "what is ur AGE," the other person answers 19, and 0fv replies "thank god" before sending explicit content captioned "this could be us."

A decoy whose whole job is to catch an adult preying on a child does not respond to "I'm 19" with relief. Relief is what you feel when you wanted the other person to be an adult. That one "thank god" is almost impossible to square with the claim that this account was bait for a predator.

The "15" only showed up later. Not at the start, the way a real decoy states it before anything happens. It surfaced mid-conversation, after the explicit material, buried in a complaint that "pedos" have unrealistic standards: "god forbid a 15 yo wants to crack a 19 yo."

The "15" arrives after the explicit exchanges were already underway, framed as an offer rather than a warning.

This is the exact reversal of how the method is supposed to work. A real decoy leads with their own age so the target has to knowingly cross the line. Dubi led with sexual content, confirmed the other user was an adult, and only mentioned a "15" later. He later told us the other user knew his account was 15 from the very first Roblox message because his decoy profile had a coded age in it. We asked him to back that up. He never did, and nothing in the documented conversation shows the other user being told an age before the explicit content was already flowing.

The other user's own reaction undercuts the claim too. When the "15" landed, they were caught off guard and started pulling away. That reaction makes no sense if they had known the age from the first message. You do not get startled by old news.

And once they tried to pull away, Dubi worked to keep them from leaving. "RELAX." "I'm not going to tell anybody." "I am a RANDOM." Then he sent pictures of Roblox condo environments.

The account talking its target out of walking away, with condo screenshots attached.

And the persona he kept up was nothing like a passive decoy. He gave it a preference and a history:

"I don't fuck like a million MAPS, if I find one I stay with them until either one of us leaves." MAP stands for minor-attracted person.

A fabricated history, offers of "files," and "do u wanna see a vid of me getting groomed by a huge Roblox dev on a condo."

He scripted the other user's lines for them:

The account literally drafts an incriminating sentence and asks if saying it "would be too far."

He solicited directly, repeatedly:

"what are some ways to turn you on besides condos."

And he pushed toward real-world contact. He sent a video of cash and offered to fly out. The other user said that would be illegal. His answer was to suggest somewhere it would not be.

"would you let me crack if I fly out," sent with a video of cash on a bed.

When told a meetup would be illegal, the account asks about Japan's age of consent and promises not to "snitch."

None of this makes the other user innocent. They stayed in the conversation after the "15" appeared, and that is on them. But the flag is about the decoy's side, and whatever the other user wanted, the decoy is never supposed to be the one moving things forward. Dubi reassured them, solicited them, wrote their lines, and offered to fly out. A real decoy waits for the target to push.

The report also documents him sending photos of a desk setup and censored explicit material with captions like "is this good enough size on soft?"

Photos sent from the 0fv account. The censored blocks are explicit material.

The real question is, why does he have explicit photos ready to send at all? This is the part no version of the honeypot story explains. Watch the work of any well-known predator hunter and you will never see one of them send a nude. The decoy does not send sexual content, full stop, because the entire point is to make the target do it. An operator who has explicit photos staged and captioned for a "sting" has already left the role of someone documenting a crime.

Whose Account Is This?

Dubi says he operated 0fv as a honeypot, so identity is not really in dispute. But it is worth showing how thin the "anonymous bait account" layer was, because it tells you how this operation was actually run.

The 0fv account asked its target to add and invite his main account, by name, twice.

"invite officialdubi to one" and "can you add officialdubi." A search of the conversation surfaces both messages.

Bloxlink lookups documented in the report resolve 0fv's Discord ID to the same linked Roblox identity associated with officialdubi, alongside an alt account network.

The report's account linkage. The 0fv profile sits on top of a Bloxlink card resolving the connected Roblox account.

The 0fv account shared two servers with the report's author, including ours.

And his main identities connect to each other through his own Discord:

The epicdubi Roblox account linked to the officialdubi Discord, through an alt account that Luna had already banned from our own server.

His Explanation

We asked Dubi to explain the conduct in the report, not the existence of a bait account. Why does a sting send pornography? Why does it confirm its target is an adult and say "thank god"? Why does it offer to fly out?

His position was that the other user was the predator and started the explicit exchanges:

His claims on May 26: the other person shared NSFW first, the document is not the full picture of a "3 month long honeypot," and nothing connects him to the clip.

Then he explained why explicit material went the other way. This is the single most important thing he has said in the entire case, so we are quoting it in full:

"to clarify why nsfw was sent back, to allow me to collect further data on these pedophiles it is imperative that i try to please these people as much as possible, i was able to collect biometric data such as the pedophiles face and other identifying features which i've documented, like i said previously that honeypot is still active to this day and i am still siphoning information from this individual successfully"

May 26, 1:47 AM. He confirms explicit material was sent from his side and frames it as investigative technique.

There is no version of the method where this sentence survives. "It is imperative that i try to please these people as much as possible" is the opposite of the one rule every predator hunter follows, which is to never give the target anything and only ever receive. The moment you start pleasing the target with explicit material to keep them talking, you are no longer documenting their behavior, you are participating in it.

He told us the screenshots lack context from a three-month operation. We have asked, repeatedly, for any of that context. Chat exports, the documentation he says he keeps, the "biometric data" he says he collected, a report filed with anyone. As of publication, we have received none of it.

After the full staff team reviewed the appeal, the flag was upheld with the wording shown at the top of this post: his actions "significantly exceeded legitimate evidence-gathering practices," and instead of documenting misconduct he "engaged with and distributed Roblox-related NSFW content." That is the whole flag. Not the clip, not his voice. The way he carried out the operation, and nothing else.

The team's final determination. The appeal was denied and the flag upheld, because we were never given a satisfactory explanation for why sharing Roblox-related NSFW content was necessary for the operation.

The Explanations Kept Growing

When someone is flagged and innocent, their story tends to get smaller and more specific over time as details firm up and evidence arrives. When someone is flagged and in the wrong, the story tends to get bigger. This case produced one of the bigger ones we have seen.

The grand jury. Dubi's central defense is that the sting's evidence is part of a New York grand jury indictment, which requires total secrecy, which is why he cannot show us anything. On May 29 he offered, unprompted, to provide specifics:

"Do you need me to provide the case number? Or the officer in charge of the case? Or the judge which is monitoring the grand jury indictment for a warrant of their electronic devices?"

May 29. He offers the case number, the officer, and the judge.

We said yes. Two days later, providing any of it had become legally impossible:

"grand jury indictments and their proceedings in the United States are legally required to remain confidential to ensure there is no jury bias. It would be contempt of court to provide you the judge residing over this proceeding as I am not allowed to legally."

May 31. The same details he volunteered are now contempt of court.

He has since told us he emailed the prosecutor personally, who advised against a court order, and offered to share screenshots of that exchange with everything except the ".gov" blurred. Layer by layer, every verifiable detail gets replaced with a story about why verification is impossible.

The legal claim is also false. Grand jury secrecy binds jurors, prosecutors, and court staff. It does not bind witnesses, who are free to share their own evidence and their own conversations.^[5]6 He sent us a government explainer on grand jury secrecy to make his point, and that same explainer spells this out. Even if his investigation is real, his chat logs and documentation are his own, and there is no rule stopping a witness from sharing them.

None of this proves there is no investigation. It might exist. Our point is simpler. The excuse he gives for not showing us anything does not hold up, and the things he says he cannot share are exactly the things that would clear him.

The grand jury sits on top of a pile of similar claims made during this case: a private lawsuit against Roblox that supposedly got his account unbanned, and a collection of screenshots and documents meant to show he had worked with law enforcement in the past. Each would be significant if true. None of them came with anything we could verify.

The retraction. This is the one worth slowing down on, because Dubi built a large part of his defense on it, and the order of events is easy to get backwards.

Here is what actually happened, start to finish. Acedia wrote the report and brought it forward in a community Dubi was part of. He was immediately criticized and dogpiled, told the conduct was a "honeypot," and told that Dubi was Rotector staff (he's not) and therefore protected. Under that pressure, he withdrew the document. On the surface, he had backed down.

That withdrawal is the "retraction" Dubi later waved at us as proof. The author apologized and took it back, so the document must be wrong.

But Acedia had withdrawn it without ever actually believing the honeypot story. His own review kept telling him otherwise. He noticed that the explicit material Dubi claimed was "off Twitter" matched Dubi's own room, camera angles, and PC setup. It came from his own desk, not from somewhere else. So he came to us, still uneasy, and told us what had happened.

Acedia's account to our team. He writes that the conduct "screamed entrapment and active involvement," that material he was told came "off Twitter" returned nothing on a reverse search and matched Dubi's own setup, and that he was "no longer confident in the original framing of the case."

The claim Acedia is referring to. Dubi told our team the explicit material was "pulled off of twitter" and "not created or manufactured by myself."

His own setup photos undercut that. One of them has a Roblox condo open on the monitor, and the report's author zoomed in to find Dubi's own account name in the player list. This was his own condo activity, not something pulled off Twitter.

The condo open on Dubi's monitor in one of the setup photos. The in-game name reads "dubi."

The photos also connect to each other. The explicit photo and the condo screenshot were taken at the same desk, down to the same mousepad and the same monitor stand in frame. This was all coming from one place, his own setup, which is not how "I pulled it off Twitter" works.

The same mousepad and monitor stand appear in both the explicit photo and the condo screenshot. Explicit areas are blacked out.

Acedia explains why he backed down: he was told Dubi "was working for rotector," so he figured Dubi was "legit." Heceta confirms Dubi was never staff, and Acedia admits he was "probably just misled."

The turning point was a single fact, and it surfaces another lie. The reason Acedia had backed down in the first place was that he had been told Dubi was Rotector staff, protected by the team, so reporting him would go nowhere. That was false. From Heceta on our team: Dubi has submitted reports as a hunter, a community role for users who send in documentation, but he has never been staff. The claim that he was, pushed on Acedia by Dubi and the friends he brought, was just leverage to get a report about himself buried. Once it fell apart, so did the pressure.

With that cleared up, Acedia stopped hedging and laid out his findings in full: the bait account did not stay passive, Dubi sent explicit material himself, and no documentation of any actual predator investigation was ever provided to him when he asked for it. He explained that the public withdrawal had been damage control, not a change of mind.

In his own words: "i didn't really believe any of his words so what i did was symbolic compliance... the evidence still stands though."

Heceta notes the documented conduct reads as involvement that would sink a real case. Acedia agrees: "no way this is honeypot," and describes being dragged into a group chat and dogpiled for saying so.

Then he put it in writing for us, through our support system, as a formal statement standing by the report:

"Much of the defense surrounding the case is relying on pressure, distortion of the facts, or confusion between what would qualify as legitimate investigative conduct and what the documented misconduct actually showed."

The author's statement to us. The findings stand. The retraction was damage control, nothing more.

So to be clear about where Acedia actually stands: he wrote the report, he stands by the report, and the only reason it ever looked retracted is that he was pressured into pretending he no longer stood by it, while being told a lie about who Dubi was. The report had even predicted this. Its closing section, written before any of the pressure started, said that any attempt to harass its author into burying it would be documented. That is exactly what happened.

The Schlep Claim

Throughout the appeal, Dubi invoked one name as his character reference: Schlep, the predator catcher whose ban from Roblox in August 2025 made international news.^[1]2

He told us everyone he consulted about his appeal, "including schlep," advised him to go public. He called Schlep his good friend who "got banned for the same reason." He warned us he would ask Schlep to make a statement about Rotector. By June 3 he was asking why people were harassing him by claiming he lied about knowing Schlep, since "the only people who know I have contact with Schlep in regards to this drama is the Rotector staff team."

June 2: "everybody i've spoken to about this appeal has told me to go public including schlep."

"my good friend schlep."

"im going to ask schlep to make a statement on rotector if stuff isn't done correctly."

June 3: upset that people say he is lying about his contact with Schlep.

This was the easiest claim in the case to check, because we were able to reach Schlep directly. On June 4, an intermediary working with our team asked him on Signal whether he knew Dubi.

Schlep's first reaction to the name: "wat."

"name sounds familiar but he doesn't work for me." Shown the officialdubi handle: "nope." Told it was also his Roblox username: "don't have him added."

The name sounded vaguely familiar to him, and that is the entire extent of the relationship. He does not work with Dubi and does not have him added anywhere. The "good friend" who advised him to go public evaporated in four messages.

By now this is a pattern, not a one-off. Take the things in this case we could actually check, and he lied about each of them. He said he never played Eviction Notice, but his own badges and his own video proved he had. He convinced Acedia he was Rotector staff, when he never was. He said his explicit photos came off Twitter, but they came from his own desk. He called Schlep a close friend, but when asked directly, Schlep did not know him. People reach for lies like these when the truth is not working for them.

It also recalibrates everything we cannot test. The Schlep claim was the one big claim in this case we could check in five minutes, and it failed. That is worth keeping in mind when weighing the grand jury, the prosecutor emails, the law-enforcement work, and the lawsuit, none of which can be checked at all.

What We Cannot Verify

For completeness, there is one more set of material, and we want to be careful with it.

Screenshots were provided to us showing an account named "Ibud" sexually harassing someone who repeatedly addresses them as "dubi" while begging them to stop. Ibud is Dubi spelled backwards.

The recipient tells the account "go away dubi" and "please stop dubi." The account responds "I love harrassing you, sexually."

We cannot verify these screenshots. We do not know who captured them, the platform metadata is not something we can independently check, and unlike everything else in this post, nothing in Dubi's own statements corroborates them. Given what we found about the clip earlier in this investigation, we hold this material to the same standard. It is unconfirmed, it should be weighed lightly, and it played no role in the flag decision. It appears here only because leaving its existence out entirely would be its own kind of editing.

Where the Sting Went Wrong

You do not need a law degree to know this sting was run wrong. Anyone who has watched a predator-catcher video on YouTube knows the shape of a real one. Predator Poachers, Schlep, JiDion, and Skeeter Jean all run on the same handful of rules. The decoy says it is a minor up front. The decoy never sends anything sexual. The decoy never makes the first move. You sit there and you let the other person walk into it on their own, and most of the time they do not. They just tell you to go away.^[3]4 The footage exists precisely because the catcher never had to do anything but wait.

Dubi's operation broke every one of those rules in a single conversation. He made first contact. He sent explicit material before any age was established. He learned the other person was an adult, was glad of it, and kept the material coming to "please" them. He played a 15-year-old only after the fact, promised not to report anything, offered "files," wrote out incriminating lines for the other user to say, and tried to arrange a meeting in a country where it would be legal.

We will be straight about what we think. We cannot see inside his head, and we are not going to pretend to. But the sting story does not survive its own evidence. A real operation leaves a trail, and after a week of being asked he produced nothing. The relief at hearing the target was an adult is not a reaction a real decoy would have, and a real decoy does not send its own explicit material. Put together, this does not read as catching a predator. It reads as taking part, with the word "sting" added after.

Dubi does not think he did anything wrong, and he believes this is simply how the job is done. But the conduct itself crossed a line. There was no need for him to send any explicit material, neither the Roblox-related NSFW nor the explicit images of himself. The flag was never a judgment on whether Dubi has reported predators, exposed predators, or even assisted in catching predators. It was a judgment on the conduct we reviewed, a sting that went past documentation and became active participation.

Acedia is the one person who got hurt for doing the right thing. He documented the conduct, and was dogpiled into a fake retraction for it.

After We Published

This post went live, and within a day he responded in a comment under the video he made about us. It announces a follow-up video about how we "deliberately manipulates, lies, and even twists evidence out of spite," and it makes a set of new claims. They can be checked the same way everything else in this case was checked.

His public comment, posted the day after this post went up. The outlined parts are the claims addressed below.

He now says he helped build Rotector. In his words, our detection is awful except for "the condo discord detection system(s) which were developed by other people anonymously including myself." This post is written by jaxron, the creator of Rotector, which makes this the easiest claim in the case to answer. I am the only developer Rotector has ever had. Every part of the system was written from the ground up by one person over the past year and a half, and there are no anonymous co-developers because there are no co-developers. Hunters like him submit reports and documentation. None of them have written any part of the system.

If he truly worked on our systems, a few questions should be easy for him. What is our technical stack? What technologies does the detection run on? What internal names do our systems go by? How does the condo detection he says he built actually work? He is welcome to answer any of these publicly. Based on how his last six lies went, we already know what answer number seven would look like.

He now says the explicit photos were made with AI. In his words, "i used ai to generate photos of myself to send to predators (i am 19), to continue and lengthen the data collection process on these individuals." Tools that generate explicit images exist, and we are not going to pretend otherwise. The problem is that this is his third story for where the same photos came from. During the appeal he wrote that the material was "pulled off of twitter and was not created or manufacured by myself." In his closing messages he wrote that he sends "fake photos of myself" and that he just pulls "porn off of adult websites." Now the photos are AI images of himself. Material that was "not created or manufactured by myself" cannot also be photos of himself that he generated, and other people's porn is not either one. The stories rule each other out, and not one of them explains why the explicit photo in the report was taken at the same desk as his condo screenshot, with the same mousepad and monitor stand in frame.

He now says Luna sent him "uncensored roblox nsfw." His comment warns readers to be careful opening "any images from luna's pov" in his transcript. Across the entire appeal, our staff sent him three images in total. Heceta sent a screenshot of the badges on his own account and a frame from his own appeal video. Luna sent exactly one image, a capture from the 0fv conversation that his flag was based on. That capture is the "NSFW" he is warning people about, and his description of it is wrong twice over. It is his own conversation, containing material he himself sent, shown to him once because he was appealing a flag about it. And it was not uncensored. The explicit content in it sits behind a solid red censor block.

The only image Luna sent him in the entire appeal, attached to her May 26 message about the 0fv conversation. The explicit content is covered by a red block.

The comment also says he is 19. Days earlier he had written that playing Eviction Notice in August 2024 "would mean i would've been like 16." Someone who was 16 in August 2024 would be 17 or 18 today, not 19. The two claims only fit together if "like 16" actually meant 17. We said earlier in this post that we cannot verify his age, and that has not changed.

He says the appeal was handled with "gross negligence." You have just read how it was handled. Every allegation we could not stand behind was removed, including the clip that started the whole case, and every quote in this post comes from his own messages. He has also now published the full appeal transcript himself. We compared his copy against ours, and nothing in it contradicts anything quoted here. The only difference is a handful of messages sent after our copy was exported. The record is public from both sides now, and readers can judge the handling for themselves.

Those final messages are worth quoting, because they describe the method in his own words:

"when i honeypot, i try to get the most amount of charges possible on an individual so they don't see the light of day again"

"if that means having to send a pedophile fake photos of myself, or faking reciprocation to the behaviors that they display then i will"

"because then they'll get way more charges than just a class E felony, which is only 4 years"

Sent after the flag was upheld. Fake photos of himself and faked reciprocation, described as how the job is done.

Sending material and faking reciprocation on purpose, to multiply the charges a target would face, is the entrapment problem from earlier in this post stated as policy. An operator who supplies the material and manufactures the escalation is not building a stronger case against a predator. He is giving the defense its closing argument.

The same messages go further:

"so unfortunately, the best way of getting these people put away is having to deal with a majority of their absolutely disgusting kinks, and then encourage it enough to where they begin to ask the honeypot for photos and such"

Encouraging a target until they ask for explicit material is the exact conduct the word entrapment describes, and here it is presented as the technique. His reason for working this way is a legal premise that his own messages disprove. In the same set of messages, he summarized the federal statute he relies on:

"Asking a minor for nudes is a federal crime under 18 U.S.C. § 2251(a), which prohibits enticing or coercing a minor to engage in sexually explicit conduct for the purpose of producing a visual depiction."

His own statute summary. The asking is the crime, with a 15-year minimum, and nothing in it requires an exchange.

Minutes later, he explained why his decoys send photos at all:

"because remember, unfortunately 'soliciting a minor' only applies when there's photos being exchanged"

His messages from the same night. The highlighted one is the premise his whole method rests on, sitting above the plan to "encourage it enough."

Do you see the difference between these two statements? In the first, he says asking a minor for photos is already a crime by itself, with a 15-year minimum. In the second, he says asking is not a crime yet, not until photos actually get exchanged, and that belief is his reason for having the decoy send photos at all. The first statement is the correct one.⁷ The asking is already the crime, so the decoy never needed to send anything back. His whole excuse for sending explicit material was the second statement, and his own first message shows it is false. This is the second time in this case that a source he cited says the opposite of what he built on it, after the grand jury explainer that confirmed witnesses are free to talk.

He is getting desperate. The threats keep coming, and his stories keep changing faster than he can keep them straight.

Where Things Stand

Here is the full accounting of how we handled this case:

• The clip: removed from his flag for review, then dropped entirely. We could not authenticate it, and we will not keep evidence we cannot stand behind.
• The voice matching: treated as moot once the clip was dropped.
• The Ibud harassment screenshots: unverified, excluded from the flag decision.
• The lies: he denied playing a game his own badges and video prove he played, he convinced Acedia he was Rotector staff when he was not, he said his explicit photos came off Twitter but they came from his own desk, and he claimed Schlep as a close friend who does not know him. After this post went live he added two more, that he helped build Rotector, which I can rule out personally as its only developer, and that Luna sent him uncensored NSFW, when the only image she sent him was censored. Every one of them checkable, every one false. None went in the flag, but together they are why we stopped extending him the benefit of the doubt.
• The sting conduct: documented in a 12-page report, corroborated by his own written admission that he sent explicit material to "please" the target, reviewed by the full staff team, and never justified despite a week of opportunity. This is the flag, and it stands.

We dropped what we could not verify, and we kept what he could not explain. The lies are not part of the flag, but they changed how we read everything else. Once someone has lied about every fact we could check, his insistence that all the rest is "doctored" and "fake" stops carrying much weight.

If Dubi wants to change the outcome, the path has not changed. We do not need the grand jury's records. We need an answer to a much smaller question: show us the operation. Decoy logs, documentation, a report filed with any platform or agency, anything that distinguishes what 0fv did from what it looks like. The things he has told us he possesses are exactly the things that would help him. We also note, in fairness, that members of our team believe he has done genuine reporting work in the past. That history is real, and it changes nothing about what is documented here, because past reports do not license present conduct.

Until then, these are the questions his appeal has never answered:

1. Why did the bait account send explicit material at all, including images of himself, when no legitimate decoy operation does that?
2. Why have there been three different stories for where those photos came from, when all of them point back to his own desk?
3. Why did he confirm the other user was 19 and respond "thank god," if the goal was to catch someone preying on a minor?
4. Why did the "15" only appear after the explicit content, with no proof the other user was ever told an age beforehand?
5. Why did an anonymous decoy ask its target to add officialdubi, his main account, by name?
6. Why did "i've never played the game eviction notice," contradicted by the badges in his own video, later become "i played this game on aug 5th 2024"?
7. Why did his offer of a case number, an officer, and a judge become "contempt of court" two days later, when grand jury secrecy does not bind a private witness sharing his own evidence?
8. Why did Acedia, the report's author, face a group chat dogpile instead of counter-evidence?
9. Why does Schlep, the "good friend" he named in his defense, not work with him or even have him added?
10. Why did defending a supposedly legitimate operation require six separate lies, each about something that could be checked?

Through all of it, he threatened to serve us papers for defamation and to make a video documenting our "malpractice." Then, despite all the insistence that he never wanted to go public, he made the video anyway. It lasted 37 minutes, most of it spent attacking the clip, the voice match, and the other things we had already set aside.

So let us be plain about it. The clip was part of the original flag, but we removed it, along with the voice, the badges, and anything else we could not prove. What the flag rests on now is one thing, taken entirely from his own words and screenshots: how he conducted the operation. Everything we could not stand behind, we dropped.

By going public himself, with the threats and the video, he is the reason the clip, the contradictions, and every gap in his story are now laid out in one place. He forced it into the open.

After a week of appeals, threats, and a 37-minute video, this is all the flag now comes down to.

If any of this feels familiar, it should. We wrote up a strikingly similar case earlier this year. Different person, different detection system, same playbook. Deny everything. Call the evidence doctored. Change the story every time the last version falls apart. Reach for borrowed authority, a grand investigation no one can see, and a cast of enemies who supposedly faked it all. It is less a defense than a script, and we have now watched two people run it almost line for line.

Cases like this are how our team gets sharper, at telling a real honeypot from someone hiding behind the word, and at holding a line under pressure and noise. We came out of it better at the job. The flag stays.