DraftKings AI scored casino bettors by expected losses, NYT finds
TL;DR
- DraftKings built a 2023 machine-learning model that scored online casino customers by expected losses per promotional dollar, internally called an 'elasticity' score.
- The company personalized roughly $400 million in promotional spending with AI in 2025 and credited data science with a 13 percent margin lift on promotion-driven sportsbook bets.
- A separate 2024 project to flag customers sliding toward gambling crisis was shelved; the responsible-gaming chief said predictive tech 'wasn't evidence-based.'
DraftKings built a machine-learning model in 2023 that scored online casino customers by how much they were expected to lose for every promotional dollar spent chasing them, according to a New York Times investigation drawn from more than 40 former employees, internal memos, presentations, Slack messages and betting records. Workers called the output an 'elasticity' score.
The weekly score processed each player's frequency, day-to-day account balance and loss-to-wager ratio, and folded in a separate estimate of whether the user was likely to quit gambling altogether. Executives told investors that data science and analytics improved margins on promotion-driven sports bets by 13 percent in 2025. The company personalized roughly $400 million in promotional spending with AI that year. The higher a player's elasticity score, the more free bets and perks the algorithm directed their way.
A former data analyst told the paper, 'We are looking for traits and features that we can target that indicate a good investment,' adding that by strict financial logic 'the best investment would be a problem gambler.' DraftKings rejected 'any implication that its marketing practices are unfair or improperly targets customers.'
A parallel project told a different story. In mid-2024, data scientist Nestor Hernandez began building a model to flag gamblers sliding toward crisis, using the same betting records to score risk before a customer cracked. DraftKings shelved it. Chief responsible gaming officer Lori Kalani told the Times that company leaders made a 'collective decision' not to use predictive technology for problem gambling because 'we evaluated that it wasn't evidence-based.'
Six of the analysts on our radar shared the piece within a day of publication.
Shared on Bluesky by 6 AI experts (top 5 by trust)
-
8/ As the @nytimes.com reported, DraftKings uses machine learning and data scientists to entice losing customers with promos to keep them going — even if those are the customers at risk of becoming addicts. The company s…
View on Bluesky → -
i do believe AI “works” and is extremely powerful. i’m sure machine learning can also be used for good. many of the uses to which it is being put seem to be both destructive and morally repugnant & maybe a little more di…
View on Bluesky → -
Tell me again how good and important A.I. is for human flourishing? Between this and the killer drones, I’m struggling to find the pro-human balance. www.nytimes.com/2026/09/19/b...
View on Bluesky → -
The difference between, say, the opioid epidemic and shoving gambling apps in front of everyone is that no-one has plausible deniability about the effects of gambling. It will absolutely ruin lives and families. They may…
View on Bluesky →
Originally reported by nytimes.com
Read the original article →Original headline: DraftKings and AI (NYT)