MLB AI Ban: Why Decision Ownership Matters in CX

What happened

Major League Baseball has moved to prohibit the use of artificial intelligence tools for in-game strategic decision-making, specifically banning managers and coaches from consulting AI-powered applications on the iPads already permitted in dugouts during play. The ruling draws a clear line between the passive, data-display functions the league had previously sanctioned and the active, AI-driven advisory role that some teams had begun exploring.

Under the new restriction, clubs may continue to use dugout tablets for approved purposes — reviewing video footage and accessing pre-authorised statistical displays — but any application that processes live game data and returns AI-generated strategic recommendations is now off-limits during a game. The ban applies to in-game use; teams are not restricted from using AI tools in pre-game preparation or broader front-office analytics work.

Why it matters

At first glance this is a sports-governance story, but it carries a sharp lesson for anyone designing services around human judgement and AI assistance. MLB's intervention reflects a growing institutional anxiety about the point at which a decision-support tool stops supporting a human decision and starts replacing it. That boundary — between augmentation and automation — is precisely the tension that CX and service-design leaders face when deploying AI in contact centres, retail environments and advisory services. Regulators and governing bodies across industries are watching how organisations draw that line, and MLB's ruling signals that "the tool was available" will not be accepted as a defence for outsourcing consequential calls to an algorithm.

From a behavioural-economics perspective, the ban also acknowledges automation bias: the well-documented tendency for people to defer to algorithmic outputs even when their own expertise should prevail. By removing the in-game AI prompt entirely, MLB is, in effect, protecting the integrity of human cognition under pressure — a consideration that customer-facing organisations rarely make explicit when rolling out AI copilots to frontline staff.

The Renascence take

Most commentary will frame this as baseball being conservative or technophobic. That misses the more interesting governance question: who is accountable for the decision, and does the tool's presence change that accountability in ways the institution is not prepared to accept?

The real issue MLB is grappling with is not AI competence — it is decision ownership. When an AI recommendation sits one tap away during a high-stakes moment, the human nominally "in charge" is no longer fully in charge, and everyone in the stadium knows it. Customer-obsessed operators should take note: deploying an AI copilot without explicitly redesigning accountability structures does not augment your frontline — it quietly erodes it. The smarter move is to define, in advance, which decisions AI informs, which it makes, and which must remain irreducibly human, then train your people to hold that line under pressure.