DraftKings AI lawsuit reaches Massachusetts attorney general

Daniel Vest received at least seventy promotional messages from DraftKings in a single month. Text messages, emails, other communications — all urging him to keep betting, during a period when he was, by his own account, losing thousands of dollars. He filed in the US District Court in Boston, and the complaint does not just allege harm. It alleges that the harm was engineered.

The lawsuit, a proposed class action, rests on a specific technical claim: that DraftKings used machine learning tools to identify customers most likely to respond to promotional incentives, including customers showing signs of problem gambling. The company has denied this directly. "DraftKings does not use AI to target customers based on losses, nor do we use AI to target customers based on indicators of potential problem gaming." That denial is clean and categorical, which makes it legally useful and also makes it the central contested fact in every proceeding that follows.

The case has moved faster than most nascent class actions because it did not originate in a plaintiff's attorney's office. It originated in a New York Times investigation, based on accounts from former DraftKings employees describing internal systems for identifying losing customers who might respond to promotional offers. The company disputed the investigation's findings. Vest's complaint uses those employee accounts as its evidentiary foundation, which means DraftKings' dispute with the Times is now, functionally, its dispute with the complaint.

Massachusetts Attorney General Andrea Campbell's office has noted that the allegations raise serious concerns about the use of technology to target vulnerable consumers. State Auditor Diana DiZoglio added that the conduct, if proven, would be unacceptable. Neither statement is a filing. Both are signals about the political environment in which DraftKings will litigate this.

The AI-targeting theory is the part of this case that extends beyond DraftKings. Sports betting operators across the industry use some version of customer segmentation. The question the Vest complaint is actually asking — and that Massachusetts officials are now asking alongside it — is whether there is a legal distinction between segmenting customers by likely spending and segmenting them by likely vulnerability. The Massachusetts consumer protection statute the complaint invokes does not address AI specifically. It addresses deceptive practices. Whether deploying an algorithmic targeting system without customer disclosure qualifies as deception under that statute is a question no Massachusetts court has answered.

A nearly identical case against DraftKings was filed in June by a Chicago-area customer. That case alleged repeated promotional contact despite visible signs of gambling disorder. Two complaints with the same theory, in different jurisdictions, drawing state attorney general attention in one of them, built on a major newspaper investigation — the litigation environment around AI-assisted sports betting marketing has changed materially in the past four months, and DraftKings is the company standing at the center of it.

The company's categorical denial is the right move, and it may well be accurate. But categorical denials in class action litigation survive only as long as discovery permits them to. The former employee accounts are already in the public record. The question a court will eventually put to DraftKings is whether its internal systems, as actually built, match the description its executives have given of them.