DraftKings Accused of Using AI to Target Problem Gamblers
A report alleges DraftKings employs AI to identify and exploit problem gamblers, pushing them to increase betting and losses through targeted promotions.
DraftKings is facing serious allegations that it uses artificial intelligence to identify and target problem gamblers, encouraging them to bet more and incur further losses. Investigations by The New York Times and ProPublica have revealed that the betting company's machine learning model analyzes extensive customer data, including betting frequency, loss amounts, and daily account balances, to predict which individuals are most likely to increase their gambling activity in response to promotional offers.
According to the reports, this AI system, reportedly developed starting in 2023, replaced a more rudimentary system. The new model assesses how much individuals might lose after receiving promotions, assigning them an "elasticity" score. This score helps DraftKings maximize its return on promotional spending by luring high-risk customers back to the platform with offers like profit boosts. Former employees involved in the system's development have corroborated these findings, detailing how the AI was designed to identify customers likely to lose more money in response to incentives.
One journalist, Jake Pearson, acted as an experimental subject, reportedly receiving an increasing number of promotions as his losses mounted. After losing nearly $1,800 on basketball bets in a single evening, he was invited to a VIP trial, eventually leading to full membership. Despite receiving automated messages to bet responsibly, Pearson found these reminders were insufficient against the product design that actively encouraged more gambling.
DraftKings has denied these accusations, asserting that its promotions are designed to reward engagement rather than exploit losses. The company claims its marketing efforts are aimed at encouraging active participation on the platform. However, the investigations also uncovered internal efforts to develop AI models with the opposite goal: to identify users whose gambling was becoming problematic and who were at risk of crisis. These projects, aimed at flagging customers who might need help with compulsive gambling, were reportedly shut down by the company.
Digital rights advocates have voiced significant concerns about how AI enables companies to process vast amounts of data for targeted advertising. The Electronic Frontier Foundation (EFF) has stated that AI "supercharges the harms of online behavioral advertising," particularly when it encourages addictive behaviors like gambling. The expansion of online betting, including casino games and prediction markets, exacerbates these concerns.
The World Health Organization estimates that 1.2% of the global adult population suffers from a gambling disorder, and that individuals betting at harmful levels account for approximately 60% of total gambling losses. The ease of mobile gambling apps, with their persistent notifications and emails, can significantly increase temptation for those struggling to abstain.
Experts advise individuals who gamble to monitor their activity, set limits, avoid chasing losses, take timeouts, and seek support if they are struggling with compulsive gambling. Resources such as the National Problem Gambling Helpline in the US and GamCare's National Gambling Helpline in the UK are available to provide assistance and connect individuals with local support services.
This situation highlights a growing ethical debate surrounding the use of AI in industries that involve potentially addictive behaviors, raising questions about corporate responsibility and the need for stronger regulatory oversight.