AIThis post was created with the assistance of artificial intelligence (AI).

TL;DR

The Electronic Frontier Foundation says DraftKings uses a machine-learning model trained on customers’ betting records to identify losing bettors and send them targeted promotions. The account is based on reporting by The New York Times and raises questions about safeguards, data use and the effect on people with gambling problems.

The Electronic Frontier Foundation says sports betting company DraftKings uses a machine-learning model trained on customers’ betting records to identify people likely to lose and then sends them targeted promotions to encourage them to bet again. The EFF’s account cites reporting by The New York Times; the report raises concerns about the use of personal data to reach customers who may be vulnerable to gambling-related harm.

Buying for a business?Offer from Amazon

Get business pricing on tech for your team

  • Business-only prices and quantity discounts
  • Tax-exempt purchasing
  • Multiple users, one account, clear invoices
As an affiliate, we earn on qualifying purchases.

According to the EFF’s summary of The New York Times report, DraftKings uses customers’ betting records to train a model that identifies bettors it expects to lose. The company then directs advertising and promotions at those customers, seeking to bring them back to the platform. The available source material does not describe the model’s inputs, accuracy or the specific promotions sent.

The EFF says DraftKings appears to rely on first-party data—information collected directly from its own users—rather than data bought from outside brokers for this model. That detail matters because restrictions focused only on the sale or sharing of third-party data would not address targeting built from a company’s own records.

The EFF characterizes the practice as a form of online behavioral advertising, in which personal data is used to tailor ads. It argues the approach can direct promotions toward people who repeatedly gamble despite harm to their finances, relationships or wellbeing. The source material does not establish how DraftKings defines a likely losing bettor or whether the people selected have been assessed for gambling problems.

At a glance
reportWhen: Reported in September 2026; the practic…
The developmentThe EFF has reported that DraftKings uses AI to identify customers likely to make losing bets and target them with promotions to return to its sportsbook.

Promotions Aimed at Losing Bettors

If the account is accurate, the practice puts a commercial incentive—encouraging further betting—alongside the risk that some targeted customers are experiencing gambling-related harm. The EFF argues that a model designed to find likely losers may also reach people it describes as problem gamblers. That overlap is a concern raised by the advocacy group; the material provided does not quantify how many targeted customers fall into that category.

The case also illustrates a gap in data policy. A company can build a targeting system from information generated within its own service, even if it does not buy outside data. The EFF says AI can process large datasets quickly and that the workings of machine-learning systems may be difficult to interpret, increasing pressure to gather data for training and refinement. These are the group’s broader concerns about ad technology, rather than independently documented details of DraftKings’ model.

Products Worth Considering

As an affiliate, we earn on qualifying purchases.

How the Targeting Model Works

The EFF’s report places DraftKings’ system within the broader practice of behavioral advertising: companies use information about people’s activity to personalize the ads they see. In this case, the reported source of the data is customers’ betting history, and the intended audience is people the model predicts are likely to lose bets.

The EFF has long called for a ban on behavioral advertising. It also warns that data gathered for ad targeting can circulate beyond its original purpose, citing sales to or use by entities including insurers, banks and government agencies. The supplied material does not say that DraftKings shared the betting data used for this model with those parties; its description instead says the company appears to use its own customer data.

The EFF points to a separate example of government interest in commercial data: it says U.S. Immigration and Customs Enforcement published a request for information earlier in 2026 about how commercial big-data and ad-tech providers could support investigations. That example is part of the group’s wider argument about data collection and does not establish a connection to DraftKings’ targeting.

“DraftKings is using its customers’ betting records to train a machine learning model to find losing gamblers.”

— Electronic Frontier Foundation

Model Details and Safeguards

The source material does not include a response from DraftKings or independently verify the EFF’s account. It also leaves open how the model identifies likely losing bettors, what data points it uses, how often it is updated, and how many customers receive the resulting promotions.

It is not clear whether DraftKings excludes customers who have self-excluded, set betting limits or otherwise sought help, or whether the company applies other safeguards before targeting them. The available account does not quantify the model’s effect on customer betting or establish how many people targeted have gambling-related problems.

Responses and Policy Debate

The next developments to watch are any response from DraftKings, further reporting on the model’s operation and evidence about the safeguards applied to customers. The supplied material does not announce a company review, regulatory action or a timetable for changes.

The EFF is using the report to press its case for a ban on behavioral advertising, arguing that limits on third-party data sales would not stop targeting based on a company’s own records. Whether policymakers take up that proposal, and whether any resulting rules address first-party data and gambling promotions, remains uncertain.

Key Questions

What does the EFF say DraftKings is doing?

The EFF says DraftKings trains a machine-learning model on customers’ betting records to identify people likely to lose, then sends them promotions intended to bring them back to bet. Its account cites The New York Times.

Does the report say DraftKings bought outside data?

No. The EFF says DraftKings appears to use first-party data collected directly from its customers for this targeting. The source material does not describe purchases of outside data for the model.

Does the report confirm that targeted customers are problem gamblers?

No. The EFF argues that people with gambling problems may be especially likely to be targeted, but the material does not establish how many customers fit that description or how DraftKings classifies them.

Has DraftKings responded or announced safeguards?

The supplied source material includes no DraftKings response and does not describe specific safeguards. It remains unclear whether the company screens out customers who have sought help or set limits.

Source: hn

You May Also Like

ByteDance Rejects Report Linking Zhou Chang’s Departure To A New Company

ByteDance has denied a report that Seed AI research leader Zhou Chang left to start a company. His current role and the report’s evidence remain unclear.

How AI Search Ads Fit Into Google’s Global Expansion

Google is expanding AI Mode ads internationally and adding ads to US follow-up questions as Search revenue growth slows.

How Choose Boring Technology Fits Into Technology Trends

An IdeaNavigator AI proposal treats the 2015 “Choose Boring Technology” essay as a test case for a role-focused technology change monitor.

Granite Switch 4.1.1 — Update Models Like You Update Software

IBM’s Granite Switch 4.1.1 updates speech recognition and a Decision adapter while keeping the base model unchanged and backward compatible.