Massachusetts Gaming Commission orders AI audit of sportsbooks

Jordan Maynard had a problem he couldn't resolve with existing rules. The Massachusetts Gaming Commission chairman announced Thursday that he had directed executive director Dean Serpa to investigate how DraftKings and every other licensed operator in the state uses artificial intelligence and machine learning — not just the one company named in a New York Times investigation, but all of them.

The Times report was the trigger. It described DraftKings using an internal measure called "elasticity" to estimate how much more a bettor might wager if offered the right promotional offer, drawing on interviews with roughly forty former employees, internal research memos, Slack messages, and customer betting records. Commissioner Paul Brodeau called the findings "troubling." Maynard stopped short of that word, but the breadth of what he ordered — every operator, not just DraftKings — signals that he read the same story and saw a systemic problem where the Times saw a single company's conduct.

Massachusetts has been here before in one specific sense: it is among the only US jurisdictions to reference AI in any of its existing gaming rules, and it commissioned a study from the University of Nevada Las Vegas on AI in gaming that recommended forming an AI taskforce. That taskforce now exists. Thursday's order is the taskforce finding its first real assignment.

The distinction that matters for anyone watching this space is between marketing AI and responsible gambling AI. The same algorithmic infrastructure that identifies a bettor's elasticity could theoretically be used to flag someone approaching problem gambling thresholds. Whether DraftKings or any other operator has built both capabilities, and whether one is being used to inform the other in the direction regulators would prefer, is precisely what Serpa has now been asked to find out.

The consensus read on this story is that it runs parallel to the New York attorney general's campaign against prediction markets — same industry, same week, same pressure. The consensus is wrong about the connection. New York is fighting a jurisdictional war over whether prediction market event contracts are gambling at all. Massachusetts is asking a different question: assuming the activity is licensed and legal, what are operators doing with the data? That is a regulatory maturity question, not a legality question, and it is considerably harder to resolve with a lawsuit.

The market angle here points toward DraftKings more than it does toward any prediction market operator. A finding that DraftKings systematically targeted losing bettors using AI would not be resolved by a CFTC preemption argument or a Supreme Court clarification of event contract jurisdiction. It would sit squarely inside Massachusetts' existing authority, and Maynard has now put that authority in motion.