344
Productivity & Workflow355
Automation & Workflow225
Software Development251
Marketing & Growth192
AI Infrastructure & MLOps175
Writing & Content Creation203
Data & Analytics142
Photography & Imaging156
Design & Creative170
Customer Support133
Sales & Outreach125
Voice & Speech135
Education & Learning131
Operations & Admin87
Former employees told the NYTimes DraftKings used machine learning to aim bonuses at customers likely to keep losing, while addiction-risk tools lagged.
In short: A report says DraftKings used machine learning to decide which customers should get betting bonuses, focusing on people likely to keep losing money.
A New York Times report says DraftKings analyzed customer betting data with machine learning, which is software that learns patterns from past behavior (like a spreadsheet that teaches itself what to look for). Former employees said the goal was to predict which gamblers were most likely to respond to promotions such as free bets and bonuses.
The reporting describes a model built in 2023 that scored customers based on how likely they were to lose money after receiving a promotional offer. Former employees said this helped the company fine-tune marketing toward bettors expected to keep losing.
The same reporting says DraftKings did not meaningfully apply similar predictive tools to identify customers at risk of developing gambling addiction. Former employees said some efforts aimed at responsible gambling were delayed or stopped.
Promotions like free bets can sound harmless, but research suggests they can push people to bet more and spend more. One study cited in the reporting found these offers increased the amount spent on bets by more than 10% and nearly cut in half the number of people who chose not to bet. Reviews of gambling research describe these promotions as inducements, meaning offers designed to encourage more gambling, including things like refunds on losing bets.
For everyday users, the concern is that the same kind of “prediction” used to send enticing offers could also be used to spot people who may be sliding into harmful behavior, but the report suggests that second use was not a priority.
Source: NYTimes