MLB's dugout iPad ban is not really a sports story. It is a preview of a conversation that is already starting in offices, courtrooms, trading floors, and insurance firms around the world: who gets to decide when AI advice becomes AI control?
Here is what happened. MLB issued iPads to all 30 teams years ago, initially just for watching game replays. The tablets had an open third section where teams could add their own apps. Some teams used it to install AI software that processed live game data and fed coaches real-time recommendations: throw a slider, pull the pitcher, shift the fielders. The Marlins pioneered calling pitches from the dugout itself in late 2025, and by this season the practice had spread to at least six other teams. The Mets reportedly paid several hundred thousand dollars for one AI program and, according to a former player now working as a broadcaster, were bragging about it early in the season.
MLB sent a private memo to general managers in June and then locked the third software section on all devices when the second half of the season started this week. As many as ten teams had been using these tools in at least one prohibited way, according to people familiar with the situation. No one was punished, because no existing rule had explicitly banned it yet.
That last sentence is the important one. The AI use was technically compliant. The league acted not because rules were broken, but because it decided human judgment needed protecting before the technology went further. One front office executive put it plainly: "Gotta stop the cheating before there's cheating now."
This is the pattern that is repeating across industries. A tool is given to workers for a limited, useful purpose. Someone figures out they can attach AI to it. The AI starts making suggestions. Suggestions become defaults. Defaults become decisions. Nobody explicitly decided to hand control to the software; it just happened gradually, through a series of small steps that each seemed reasonable.
The Mets' case is also a useful warning about cost versus outcome. The team reportedly spent several hundred thousand dollars on AI-assisted pitch selection and still finished near the bottom of the league. AI can process more data than any human coach, but it works from historical patterns. A baseball game, like most real business situations, also involves momentum, fatigue, weather, and the specific psychology of the person standing at the plate in that moment. The Marlins, who started the whole trend without elaborate AI software, at least had the courage to test an idea and measure it honestly.
For business operators outside sports, the relevant question is not whether AI tools are useful for preparation and analysis. They clearly are. The question is whether your organization has drawn a line between AI that informs decisions and AI that replaces them, and whether your staff actually knows where that line sits.
MLB drew its line by locking a tab on a tablet. Most organizations will need to do it more deliberately, because the tools are already inside the building and the rules that govern them are still being written.