NYT Investigation: DraftKings Used AI to Target the Gamblers Most Likely to Lose

kyliebytes · x · 2026-09-22

A New York Times investigation reveals that DraftKings built a machine learning model in 2023 on customer betting records to identify users most likely to respond to promotions by gambling — and losing — then targeted them with free bets and bonuses, spending hundreds of millions annually on such incentives. Former analyst Jayden Butts, who tested the model, worried these same users were prone to addiction: by strict financial logic, "the best investment would be a problem gambler." When employees later developed a model assigning "risk scores" based on users' likelihood of developing compulsive behaviors, the company sidelined it. Based on internal documents and interviews with dozens of former employees.

Original post →

More from AGI Musings

AGI Musings channel →