When the Vendor Films the Cheat: ACR’s TrenirusGTO Video and the Real Integrity Gap

On September 17, 2026, a video hit X showing an ACR Poker cash table sitting next to a GTO overlay that appeared to spit frequencies and action advice in real time. It was not a sting. TrenirusGTO posted it as a product promo. firenkes and others amplified it; views cleared 110,000 the same day. Chris Moneymaker did not hedge: this is not study, it is blatant cheating.

ACR published a long statement within hours. RTA is strictly prohibited. An investigation is open. Confirmed users get a permanent ban, funds confiscated, money returned under game-integrity procedure. CEO Phil Nagy was shorter: ACR is "the last place you want to use RTA on." Catch you, ban you for life.

Those lines are correct. They are also not the story. The crack worth writing about is this: a software vendor was willing to advertise "open a solver against live humans" as a feature. Real-time assistance has moved from a basement product to a public acquisition funnel.

Pin the timeline before the outrage

Only what is public. No reconstructed streets, no invented hole cards.

Beat	What is public	What kind of source
Origin	TrenirusGTO posted the clip to promote the tool; the player's identity is redacted	Poker.org
Spread	firenkes: this is RTA — a solver running during the hand against real people; if rooms do not block it, cheating is legalized	Same; community accusation, not a court finding
ACR first statement	The video shows TrenirusGTO used in real time alongside ACR; RTA is banned; Security and Game Integrity is identifying the account, the scope, and any related accounts	ACR statement, reprinted in full by Gambling911
CEO	Phil Nagy: last place you want to use RTA; lifetime ban if caught	X, via Poker.org
Ambassador	Moneymaker named TrenirusGTO: not improving your game; blatant cheating; funds confiscated	Same
Pros	Patrick "Pads" Leonard: "very scary," and this is less than 1% of the RTA talk regs have had all week. Matt Berkey: AI makes these tools cheap; the endgame he struggles to unsee is operators and cheats taking the money	Same; player commentary, not a platform verdict
Policy already on the books	ACR Security and Game Integrity 4.2.4 already bans programs that output live ICM / Nash / EV / GTO advice on the current hand. 6.5 treats ghosting as RTA	ACR policy

Gambling911 later the same day quoted a follow-up: the username on screen was UGLYRETD, created 8/2/2026, deposited $14.13, played one day — 65 hands at $0.05/$0.10, 27 at $0.02/$0.05, 19 at $0.01/$0.02, net -$13.12, permanently banned; seven players who lost to that account would be refunded. Treat those figures as ACR's own disclosure, as relayed by the trade press, until they are re-checked on ACR's channel.

Even if every number is exact, they answer a narrow question: was the promo shot on a microstakes demo account? They do not answer Pads.

The line was never the logo. It was whether the hand was still live

RTA is not a brand. It is a class of pipes. If a tool compresses, replaces, or removes your decision while the hand is live, it is cheating. Preflop charts, river solvers, ICM apps, a second-screen frequency overlay, a coach yelling jam on Discord — same class.

Legal study versus illegal assistance is timing, not chrome:

	Study hall (legal)	Exam hall (RTA)
Client	Closed, or no pending action	Open, a live pot waiting on you
Solver / GTO trainer	Historical hands, ranges, equity review	Current board, stacks, and action fed in; you click the output
Overlay	None	Frequencies and recommended sizing sitting on the ACR window
Result	You might get better	The field is playing an engine

If TrenirusGTO only ingested finished hands after the session, it would be a trainer. The usage alleged in the clip is GTO output stacked in front of a live ACR decision. ACR's statement is explicit: they are not investigating whether that is allowed. It is not. Section 4.2.4 already said so.

This is not a gray rule. Someone filmed the closed-book exam as an open-book ad.

For the industry definition, see what RTA is and why rooms ban it. For the capture-relay-solve stack, see Super RTA. For GGPoker's air-gap rule while the client is open, see can you use RTA on GGPoker.

The phase change: the vendor submitted the evidence

RTA at online tables is not new. Kruse's two-machine setup in 2020, the 2025 screenshot of ACR ambassador Nacho Barbero with GTO Wizard on screen — same incentive. Once solver latency drops to a few seconds, someone will turn "look it up" into "ask it while you play."

Three things are different.

Acquisition went public. RTA vendors used to pretend this was underground. This time the product account posted a suspected live ACR overlay as a demo. Poker.org had it right: the software company is openly promoting a prohibited use of its own product.

Detection got hijacked by advertising. Rooms catch who is willing to be seen. If UGLYRETD is real, it is a $14 deposit that lost $13 — perfect for a commercial, useless as a map of high-stakes risk. The Super RTA that actually drains liquidity does not ship an 110k-view trailer. Pads's "less than 1%" is the asymmetry: the demo account is visible; the silent pipes on mid- and high-stakes are not.

AI crushed the supply curve. Berkey does not need a conspiracy. Wiring "read the table, dump frequencies, wait for a human click" is no longer "run a solver farm." It is "ship a product." The ending he cannot unsee is rake on one side, cheats on the other, and the people who get paid to think in ranges, blockers, and sizing getting flattened in between.

Recreational games die first. firenkes was not being dramatic. Skill collapses into whoever has the overlay open. Fish who believe they are not playing a person do not learn GTO. They leave. Regulars do not lose winrate to better-studied humans. They leak it to software.

ACR's statement was fast. That does not erase the file

Keep two facts apart.

On September 17 the copy was cleaner than January 2025's "Nacho is a goofball and we love him." Overlay was not recast as personality. The rule, the penalty (permanent ban + confiscation + redistribution), the GTO Wizard integrity partnership, and the Final Table Replayer — cards-up after the tournament ends — were stated as opposites of live assistance. They should stay opposites.

The other fact: ACR's integrity reputation did not start this morning. Ambassador screens with GTO tools, a history of bot and collusion allegations, a US-facing room outside a hard regulatory perimeter — all of that discounts a "lifetime ban" press hit. Players will ask the operational questions: can process scanning see an overlay? Can behavior flags catch a hundred-thousand-hand mixer that hugs solver output, not a 65-hand micro account? Are refunds seven microstakes losers, or the people who have been farmed for months at real stakes?

Nagy's line — last place you want to use RTA — is earned only by detections that do not require the vendor to file the evidence. Catching the demo account is the easiest exam on the sheet.

Moneymaker's anger maps onto the marketing lie. Calling RTA "study" steals the subject. Improvement is you recomputing SPR and explaining why 75% pot on this runout puts maximum leverage on the opponent's thin made hands. Improvement is not outsourcing EV to a second monitor while the time bank is still lit. Confiscation is not revenge. It is returning stolen equity — if the room can actually find the pots.

A microstakes demo does not close the structural case

If the follow-up is accurate, UGLYRETD is almost certainly a filming mule: one day, three micro limits, down $13. Using that as proof that "ACR does not have an RTA problem" inverts the evidence.

• Ads pick microstakes because expected loss if caught is tiny and the screenshot still proves "it hooks to ACR."
• A losing day does not prove a clean tool. A hundred-hand sample is variance. RTA's damage is long-run, repeatable proximity to optimal, not a single session ledger.
• Refunding seven players is the correct minimum. It is not an ecology report.

Three metrics matter: related accounts, the same stack appearing at mid/high stakes, and proactive catches that did not start on X. ACR will not publish detection methods. Fair. What players can demand is output: ban counts, confiscation totals, number of players made whole — numbers, not "we take this seriously."

The line you can still draw yourself

You do not run the room's arms race. You do run your tool chain.

1. Close the client before the solver. Alt-Tab on the current river and dropping a hand history into a trainer after the session are not the same act in policy or in ethics.
2. Ignore "I only look at frequencies, I still click." ACR's language is any real-time advice that reduces, influences, or removes your own decision. Whose finger is on the mouse does not change the pipe.
3. Report with evidence, not a posse. This clip spread because the picture was the evidence. "This person plays too GTO," with no overlay and no timestamp, is usually not enough.
4. Leave learning for after the session. Review, range comparison, equity, tagging mistakes — that is how the game is actually learned. A live overlay burns that path down.

PokerJudge sits on the study-hall side of that line: hand review, AI coach, and calculators after the session ends, never feeding advice into a live table. That is not a slogan. It is the condition for staying in this industry.

Integrity trends when cheating agrees to be filmed. TrenirusGTO's ad did one thing the vendor may not have priced in: it turned RTA from a black box you cannot explain to a recreational player into a split-screen anyone can read. The clip will age. The rule will not. While the hand is alive, the decision has to be yours.

FAQ

Is TrenirusGTO RTA?

Depends when you use it. Training on finished hands after the session is study. The usage alleged in the video — GTO advice during a live ACR pot — is RTA under ACR 4.2.4 and under every major room's closed-book rule. ACR already said they are not investigating whether that use is allowed.

The account was a microstakes loser. Why write this?

Because the clip is customer acquisition, not a P&L. A $14 deposit is enough to shoot a commercial. It cannot represent, and cannot rule out, silent RTA at mid and high stakes. Leonard's "less than 1%" is closer to the real size of the problem than the account ledger.

Is post-session solver or AI-coach review cheating?

Major rooms ban external decision input while a hand is live. Closed client, official hand histories, review — that is study. The same software on a second screen steering the current street is cheating. Timing, not the logo.

Nagy said lifetime bans. Is this closed?

For the account in the video, if the follow-up holds, yes: permanent ban and refunds queued. For "is there RTA on ACR," no. Catching a vendor who mailed in the tape only proves the room moves when the camera is on.