DraftKings faces second artificial intelligence targeting lawsuit

DraftKings faces second AI targeting lawsuit as Massachusetts joins scrutiny

Daniel Vest spent money he could not recover, and he kept a record. Over one month, he counted at least 70 promotional messages from DraftKings — emails, texts, other push communications — each one nudging him back to the platform. The West Virginia resident has now filed a proposed class action in US District Court in Boston, alleging that the company used artificial intelligence to identify him and others like him as high-value targets: customers whose losses made them likely to respond to promotional offers, including customers showing signs of problem gambling.

DraftKings has denied this. The company's public statement is direct: it does not use AI to target customers based on losses, and it does not use AI to target customers based on indicators of potential problem gambling. The denial is categorical. The lawsuit, which draws heavily on a New York Times investigation that quoted former DraftKings employees describing machine learning systems for identifying losing gamblers, is in its earliest stages. Nothing in the complaint has been proven.

What has shifted is the institutional response. Massachusetts Attorney General Andrea Campbell's office said the allegations raise serious concerns about the use of technology to target vulnerable consumers. State Auditor Diana DiZoglio went further, calling the conduct unacceptable if true. Two senior state officials engaged, on the record, within the same news cycle as the filing. That is not a routine response to an early-stage civil complaint.

A Chicago-area plaintiff filed a similar suit in June, alleging that DraftKings repeatedly offered him gambling incentives despite visible signs of a gambling disorder. The June case is still pending. Vest's filing means there are now two active federal complaints making substantially the same allegation against the same company, in different jurisdictions, in the same calendar year. The pattern matters more than any single filing.

I have seen this sequence before with consumer-facing platforms in regulated industries. The first lawsuit is a test. The second one tells you whether the first found anything. When the second complaint arrives before the first has had time to settle or be dismissed, and when it draws on the same investigative reporting and the same category of former-employee testimony, the company's legal exposure is not additive — it is multiplicative. Regulators who were watching the first case now have a reason to coordinate.

The AI element is what makes this hard to resolve quietly. DraftKings can dispute the characterisation of its marketing systems, but it cannot unring the bell on the underlying question: what did its algorithms optimize for, and who were the highest-scoring users? Discovery in either case will reach for internal model documentation, training data, and the targeting criteria that defined a "responsive" customer. That material, if it exists and if it reaches the public record, does not stay confined to one courthouse.

Massachusetts officials are not the CFTC. They are not arguing about swap definitions or federal preemption. They are arguing about consumer protection law in a state that already has a functioning sports betting regulatory framework. That is a narrower, cleaner argument — and for DraftKings, a harder one to deflect on jurisdictional grounds.