The World Series of Poker has introduced a new AI feature that estimates when players are bluffing by analyzing their body language, giving TV viewers new insight into the action at the table.
The technology debuted as the WSOP returned to ESPN for the first time since 2021, with broadcasts produced by Peyton Manning's Omaha Productions.
Developed by AI engineer Luke Geel, the system was trained over six months using historical poker footage and tournament results. Instead of relying only on betting patterns, it analyzes visual cues such as posture, facial expressions, and blink rate to estimate whether a player is bluffing or holding a strong hand.
To protect the integrity of the competition, the AI's predictions are only shown after a hand has ended, preventing players from gaining any live advantage.
Geel said the project was far more complex than expected and noted that some poker players have already asked whether the technology could help them study opponents or identify their own unconscious tells.
While experienced professionals are likely better at concealing physical tells, Geel believes future AI systems could become even more accurate by incorporating additional signals, such as heart rate or subtle skin color changes, if more advanced camera technology becomes available.