TL;DR
The Electronic Frontier Foundation says DraftKings uses a machine learning model trained on customers’ betting records to identify losing gamblers and target them with promotions. The claims, which EFF attributes in part to a New York Times report, raise questions about how gambling platforms use customer data and whether those identified as vulnerable are being encouraged to keep betting.
The Electronic Frontier Foundation says DraftKings uses a machine learning model trained on customer betting records to identify people likely to place losing bets, then sends some of them targeted promotions encouraging them to return. The allegations, which EFF says are based on reporting by The New York Times, raise concerns about whether gambling data is being used to re-engage customers who may be experiencing harm.
Get smart everyday buys delivered free with Prime
- Fast, free delivery on millions of items
- Prime Video, Amazon Music and more included
- Member-only deals all year
According to EFF’s account of The New York Times report, DraftKings uses its customers’ betting histories to train a model that finds gamblers likely to lose. The company then targets those customers with advertising and promotions intended to bring them back to the site to place more bets. The supplied material does not detail the model’s inputs, how many customers it identifies, or how the promotions are delivered.
EFF argues that the approach aligns the platform’s marketing with its financial interest in customer losses. It says people it describes as “problem gamblers” may be especially likely to be included: the term refers to people who continue gambling despite harm to their well-being, finances or relationships. That is EFF’s characterization of the potential impact; the source material does not provide figures showing how many affected DraftKings customers meet that description.
EFF says DraftKings appears to rely on first-party data—information collected directly from its users—rather than data bought from outside brokers to power this targeting. The organization uses the example to argue that limits focused only on third-party data sales would not address promotions built from a company’s own customer records.
How Targeted Betting Promotions Raise Risk
The concern is that the system may use signals of past losses to decide who receives offers to bet again. If EFF’s description is accurate, the same activity that can indicate a customer is losing money could help the company select that customer for marketing. That creates a potential conflict between customer protection and the commercial value of repeat betting.
The report also points to a wider data issue. Targeted advertising depends on collecting and analyzing personal information; machine learning can process large records to sort users into groups for outreach. EFF argues that this can intensify existing advertising harms and make the decisions harder for customers to see or understand. The supplied account does not establish whether DraftKings’ model is opaque in practice or what safeguards, if any, are applied to prevent harmful targeting.
Products Worth Considering
Apple Mac mini Desktop Computer with M6 Chip (2026)
Logitech MX Master 4 for Mac Wireless Ergonomic Mouse, White Silver
Samsung Odyssey OLED G5 27-Inch QHD Gaming Monitor
As an affiliate, we earn on qualifying purchases.
The Data Behind DraftKings’ Promotions
EFF describes behavioral advertising as personalizing ads based on information collected about a person. In this case, the reported data source is customers’ own betting records. EFF says that detail matters because a policy focused solely on companies sharing or selling data to third parties would not necessarily prevent a platform from analyzing its own records to choose whom to promote.
EFF connects the issue to the broader ad technology market, saying data gathered for targeted advertising can also be shared with or sold to other organizations. Its report cites government interest in commercial data and ad technology, including an Immigration and Customs Enforcement request for information earlier in 2026. That wider concern is separate from the specific DraftKings allegation: the supplied material does not say that DraftKings data was shared with ICE or any other government agency.
EFF has long advocated banning behavioral advertising. It also points readers to its Surveillance Self-Defense resources for guidance on limiting data exposure online and through mobile apps. Those are the organization’s policy position and resources, rather than findings about DraftKings’ practices.
“DraftKings is using its customers’ betting records to train a machine learning model to find losing gamblers.”
— Electronic Frontier Foundation, describing the reported targeting
What the Model and Safeguards Reveal
The supplied report does not state how DraftKings defines a “losing” gambler, how accurately the model predicts future losses, or how many customers receive these promotions. It also does not describe the specific offers, whether customers can opt out of this targeting, or whether DraftKings excludes people identified as at risk. The account attributes the core allegation to The New York Times but does not include a direct response from DraftKings or independent technical evidence about the model. Those details remain unconfirmed in the material provided.
Questions for DraftKings and Regulators
Further reporting or a response from DraftKings could clarify what data the model uses, how the company defines and evaluates likely losses, and what protections apply to customers who may be at risk. Any regulatory or policy response would also depend on the rules governing gambling promotions and the use of customer data. EFF is urging policymakers to ban behavioral advertising; the supplied material does not identify a specific proposal or pending decision tied to DraftKings.
Key Questions
What does EFF say DraftKings is doing?
EFF says DraftKings trains a machine learning model on customer betting records to identify losing gamblers and sends some of them promotions intended to bring them back to bet.
Has DraftKings confirmed the model or its purpose?
The supplied material does not include a statement from DraftKings. EFF attributes its account to The New York Times, and details such as the model’s inputs and safeguards remain unclear.
What does “first-party data” mean here?
It means data DraftKings collects directly from its customers, such as their betting records. EFF says the targeting appears to use this data rather than information bought from outside sources.
How many customers are affected?
The supplied source material gives no number for how many customers the model identifies or how many receive targeted promotions.
Source: hn
