A DraftKings developed artificial intelligence models to identify customers most likely to respond to promotions by betting and losing more money, according to an investigation published by the The New York Times. The company disputes that its offers are determined by users' losses.
The system began to be developed in 2023 with data from the bettors themselves and sought to estimate how much each customer could generate in additional losses after receiving a promotion.
The model considered factors such as gambling frequency, account balance, amounts wagered and lost, and the probability of abandoning the platform. In one of the tests, about 5,000 online casino players were analyzed.
Former employees told the NYT that DraftKings continued to improve data science techniques to target incentives at customers deemed more profitable. Similar models were also applied to sports betting.
DraftKings denies that losses determine offers
The company said its promotions are targeted based on continued use and customer engagement, not on losses.
DraftKings' public documents show that the company uses data science, segmentation and metrics such as customer lifetime value, retention and return on investment to define promotional strategies.
The company reported US$ 6.05 billion in revenue in 2025, up 27% year over year.
The investigation also pointed out that internal projects aimed at predicting players at risk of developing gambling problems were halted or not implemented. DraftKings said it did not find sufficient evidence of effectiveness in those models and says it uses other behavioral indicators to identify risk situations.
