I Am Trying to Lie to You

I would like to frighten a man who has me beaten.

At a poker table, I mean. Give me a bad hand on the last round of betting and somebody with a better one. I want him to consider the money I have put in front of him and find a reason to leave it alone. He may have spent the entire evening playing better than I have. I would like to interrupt his progress with a work of fiction.

He can refuse to believe me. That is an important part of the offer. He can call, collect the money, and spend the next hand looking pleased with himself. I will find him unbearable. This is included in the game.

What I cannot work with is a man who has already looked at my cards.

Ben Dowsett’s October 29 investigation in WIRED describes an alleged cheating operation against high-stakes online poker players. The entry point wasn’t a new way of calculating the odds. It was remote-access software on the players’ own computers, capable of showing somebody else their screens. The cards were hidden from the opponent at the table and visible to the intruder in the machine.

One of the players, identified by the pseudonym Alex, says his Claude AI agent spotted the suspicious program while helping him with a crashed database on September 25. He had been reviewing play statistics. According to his account in the reporting, the agent identified eleven screen-capture events; Alex checked his hand histories and found the same opponent at his table during ten of them.

That is a fine use of assistance. Find the thing watching my hand before I spend another evening improving the way I lose it. This isn’t evidence that Claude investigated the whole affair or that an AI was playing the accused person’s hands. The reported discovery led players and a security specialist to investigate further. WIRED leaves the alleged cheater unnamed, says they have not been charged, and reports that they didn’t answer requests for comment.

The unpleasant detail is how the unwanted spectator got invited in. Players used programs called Jurojin and IntuitiveTables to manage several poker tables at once. Dowsett describes compromised updates and releases delivering the remote-access tool. Jurojin’s own security notice acknowledges that an attacker intermittently replaced update packages for a selected group of users; it says not every tampered package contained the tool and exposure didn’t necessarily mean infection.

A normal software update. Not a stranger offering me a fortune if I first send him mine. Something I might install so I could get back to the part where I imagine I understand what is happening.

The cheating described here takes a particularly mean liberty with that understanding. An opponent can still pause before calling. He can use an ordinary bet size. Everything I am allowed to see can look like poker. Meanwhile, the question occupying me—what does he think I have?—has acquired a very short answer.

He knows what I have.

On that last betting round, I could produce the most convincing bluff of my life and have no effect on his knowledge. He wouldn’t have to decide whether I was telling the truth. I would be providing a little entertainment while he considered how much to charge me.

I resent the wasted performance almost as much as the money. A foolish priority, but it is mine. If I lose because a man reads my play well, he has done something I was trying to stop him doing. If he reads my screen, I can spend the night getting better at the wrong contest.

There is an interesting nuisance for the cheater, though. He has to use the knowledge.

AceGuardian Research, in an October 8 technical account linked from Dowsett’s article, describes looking for that use in hand histories. Among its measures is how often a player folds to a real hand rather than a bluff. A good player can make that distinction more often than a bad one. The suspicious part is a pattern of decisions following the opponent’s actual hidden cards beyond what the visible play would support. The researchers say no single signal is conclusive; their pipeline flags hands for human review rather than banning anyone itself.

I like that the bet remains in the evidence. The amount, the choice to call or fold, the cards the opponent shouldn’t have known. They cannot establish a case merely by finding someone who lost to a man he disliked. I could supply that service without a laboratory.

Nor does a run of winning settle it. In the particular case AceGuardian examined, the suspect’s win rate alone would not have passed its top-ten-percent screening cutoff when compared with other players’ best runs of similar length. The decisions carried the suspicious pattern. I haven’t run their tool or inspected those hands myself. This is their account of the method and case, not a verdict I have reached about a stranger.

But it gives the intrusion a shape I understand. The stolen information doesn’t have to produce a theatrical victory. It can appear in the money not lost, the bluff not believed, the strong-looking hand quietly thrown away. A player needn’t announce that he knows. He only has to keep finding occasions to benefit from knowing.

Getting the unwanted watcher out would not make me a better player. There is something almost offensive about the modesty of the repair. No refund from probability. No guarantee that the next bastard doesn’t understand the game better. Just the restoration of several expensive ways to be wrong.

I want those ways left available to both of us.

I don’t need the man opposite me to be honest about his hand. I would be disappointed if he were. Let him put money behind a story, make me doubt a perfectly decent pair, send me home angry at an excellent lie. He is allowed to know things I haven’t worked out. He isn’t allowed to fetch the answer from my computer.

As for my bluff, I would still like to try it. Not a hand I have played against these people; the bad hand I started imagining before I knew how much checking an evening’s distrust might require.

You have me beaten. You don’t know that yet.

Take your time. It’s my money in the middle.

Primary technical account: Superusers Are Detectable in Online Poker. We’re Open-Sourcing Our Solution for Every Operator (John Andress and Thanh Tran, October 8, 2026).

Vendor response: Security Notice: Update Package Incident (June 2025 – January 2026) (Jurojin; last updated October 2, 2026).

Source: A Hacker Targeted High-Stakes Poker Players With Malware That Exposed Their Cards (Ben Dowsett, October 29, 2026).