Major League Baseball has told its teams to keep the robots out of the dugout. A June 11 memo from the commissioner’s office banned clubs from using dugout iPads to run AI-powered applications that generate in-game strategy recommendations, cutting off a practice the league says had quietly spread to as many as a third of all MLB organizations.
The Athletic first reported the memo, which targets custom apps that push tablets beyond video review and statistical reference into live decision-making — substitutions, pitch calls, defensive shifts. The ban does not strip iPads from the dugout entirely. Teams can still use them to pull up video, look at numbers, and review sequences between innings. What they can no longer do is let a generative AI tool hand them a recommendation and act on it while the game is still being played.
No clubs will face punishment. MLB’s review determined all 30 organizations were already compliant with the new rules by the time the memo circulated, according to The Athletic’s sources. The league framed the move as preventive — shutting the door before AI-generated strategy becomes a standard part of in-game management.
The policy lands mid-season, which is unusual, but it fits a pattern. MLB tightened restrictions on dugout technology after a sign-stealing scandal surfaced in 2021, when teams were caught using electronics to decode opponents’ signals in real time. iPad use during games faced new limits in the aftermath. Teams later pushed for more lenience, and the commissioner’s office loosened some restrictions. Now it is pulling back again, this time aimed not at cheating but at automation.
The distinction the league is drawing matters. MLB is not retreating from technology. The Automated Ball-Strike system, which uses tracking cameras to call balls and strikes, continues to roll out across spring training and select minor-league venues. Statcast feeds reams of biometric and tracking data into every front office in the sport. The line the commissioner’s office has drawn is specific: tools that process data are fine. Tools that replace a coach’s judgment during a live game are not.
That boundary touches a tension baseball has been chewing on for a decade. Front offices already build lineups, shift defenses, and sequence pitchers using models that crunch spin rate, launch angle, and exit velocity. The dugout, though, has remained the last domain where a manager reads a hitter’s body language, trusts a catcher’s instinct, or goes with a gut feeling that the numbers say is wrong. Letting an app whisper “bring in the lefty” into a coach’s ear in real time would erase that gap between the spreadsheet and the dirt.
Baseball decisions are not made from numbers alone. A pitcher’s fatigue, a hitter’s frustration, the weather rolling in — all of it complicates the mathematically optimal answer. The league’s position, as laid out in the memo and reported by The Athletic, is that those variables belong to the people in uniform, not the software in their hands.
Tablets stay. The chatbots are done. And for now, the third base coach still has to make his own bad decisions the old-fashioned way.
Sources:
- Yahoo Sports — “MLB Wants Coaches, Not Chatbots, Calling the Shots”
- The Athletic (referenced in Yahoo Sports report; original article not independently accessible)