Massachusetts opens an AI targeting inquiry into DraftKings

Jordan Maynard did not call it an investigation. He called it an evaluation, and the word choice matters. The Massachusetts Gaming Commission chairman announced Thursday that his staff would examine how DraftKings and other licensed sports betting operators use artificial intelligence to identify and market to bettors, following a New York Times report that described the company deploying a metric called "elasticity" — an internal measure of how much more a given gambler might wager if offered the right promotional nudge.

The Times built its account from interviews with roughly forty former DraftKings employees, internal research memos, Slack messages, and customer betting records. What it described was not a glitch or a rogue experiment. It was a systematic practice: using machine learning to find the bettors most likely to lose and concentrating promotional spend on them.

Maynard framed the response carefully. He wants staff to engage with DraftKings first, understand the specifics, then look at how all operators in the state are using the technology. Any policy action, he said, comes after that. Commissioner Paul Brodeau called the report troubling but added the caveat that customer-facing AI optimization is not unique to gambling — it describes most of consumer retail. That is true, and it is also the kind of observation that tends to slow regulatory responses rather than sharpen them.

Massachusetts has been ahead of most states on this particular question. The Gaming Commission commissioned a study from UNLV's International Gaming Institute focused specifically on AI in marketing and player acquisition, and the recommendations from that study — released last November — included forming an AI taskforce. The commission formed one. What it has not yet done is establish enforceable standards for what operators may and may not optimize for.

That is where I think the consensus read on this story misses something. Most coverage treats Thursday's announcement as a moment of regulatory momentum — a capable commission taking a credible step. I read it as the opposite. Maynard's framing gives DraftKings the first conversation, lets the company explain itself, and defers any policy action to a future determination that "may or may not" be appropriate. That is not enforcement posture. That is the structure of a review designed to take time.

I am adjusting for my own tendency to weight the downside scenario, but even after that adjustment, the mechanism here is soft. The commission has rules that reference AI, but no rules that prohibit targeting bettors on the basis of predicted loss elasticity. Without that specific prohibition, the gap between what DraftKings allegedly did and what Massachusetts can actually sanction is wider than the announcement suggests.

The harder question — one that neither the commission nor the Times report resolves — is whether optimizing for bettor engagement is categorically different from optimizing for bettor harm, and whether any regulator has the legal architecture to enforce that distinction. Massachusetts may be the state most likely to find out, but Thursday's announcement does not yet indicate it is trying.