ESPN introduced an AI system that identifies potential poker bluffs by analyzing players' body language during the World Series of Poker Main Event.
Developed over six months by AI engineer Luke Geel, the computer vision tool tracks blink rate, eye gaze, posture, and chip handling. It compares these behaviors with past hand outcomes to assess hand strength. The technology is only applied to players already eliminated from the tournament.
Geel, who also creates AI for the US Air Force, revealed the debut on LinkedIn, noting the project was more challenging than anticipated. Final table coverage featuring this AI will air during the first week of August.
Responses from players and viewers have been mixed. Some accused the system of cheating, while others expressed interest in using it to study their own footage and learn how to mask tells. Skeptics question the AI's purpose since live broadcasts show players' hole cards anyway.
This approach could eventually extend beyond poker to analyze negotiators or job candidates, raising concerns about the darker implications of AI technologies. Whether the feature will continue next season may depend on audience acceptance.
