DraftKings AI lawsuit reaches federal court as class action bid widens

A West Virginia man says he received at least seventy promotional messages in a single month. That number is the spine of the federal complaint now filed in Massachusetts, and it is the kind of detail that tends to survive the early rounds of litigation, because it is specific, it is countable, and it is the sort of thing a company's own systems would have logged.

The lawsuit, filed as a proposed class action, alleges that DraftKings used artificial intelligence to identify customers classified internally as likely to continue betting after receiving promotional contact. The company's term for this category, according to former employees cited in a New York Times investigation that preceded the legal filing, was "elasticity." The word matters, because it reframes what a promotion is. A bonus offered to a casual customer to keep them engaged is marketing. A bonus offered to someone a system has identified as financially responsive to stimulation is something regulators in several states are now trying to name.

DraftKings has denied the characterization in full. The company's position is that the Times reporting relied on disgruntled former employees and that its promotional practices are standard across the industry. That defense may be accurate and may also be precisely the problem. If the targeting described is standard, the question becomes whether the standard is lawful, and that is a question a class action is well positioned to force.

Maine's Gambling Control Unit has said it is monitoring the situation without yet finding a Maine-specific rule that applies. This is the pattern that precedes new rules — regulators acknowledge the gap before they fill it. Massachusetts has opened a formal investigation. The federal lawsuit is now in the same jurisdiction.

The consensus read on this story is that it is primarily a reputational risk for DraftKings, manageable through denials and eventual settlement. That reading underweights the structural exposure. The allegation is not that DraftKings marketed aggressively — that is legal and common. The allegation is that the targeting criteria included behavioral signals associated with continued loss-chasing, which, if demonstrated, crosses from marketing into something that gambling harm frameworks are specifically designed to address. Proving it requires discovery into proprietary systems, and that discovery, if granted, would expose methodology that DraftKings has every incentive to keep sealed.

There is a version of this case that settles quietly and changes nothing. There is another version in which the discovery process becomes the story, and what emerges from it shapes how every state with an open AI monitoring question decides to act. Maine's Gambling Control Unit said explicitly that no current rule covers this conduct. Legislatures tend to read court filings.