DraftKings Is Using AI to Behaviorally Target Chronic Gamblers

Targeted gambling promotions based on AI-driven predictions of losing bets will now be more effective and frequent.

DraftKings is employing artificial intelligence to identify and target customers who are most likely to lose bets and respond to gambling promotions. The company trains machine learning models on betting records to find these individuals, then sends them personalized advertisements designed to encourage further betting. This practice leverages what is known as online behavioral advertising, where ads are tailored based on collected user data. The AI-driven approach aims to re-engage problem gamblers, who are profitable for DraftKings due to their propensity for continued betting despite negative consequences. The use of AI allows for faster processing of large datasets, amplifying the effectiveness and potential harms of this targeted advertising strategy. DraftKings reportedly uses only first-party data, collected directly from users, for its AI models.

The integration of AI into behavioral advertising significantly escalates its potential for harm by enabling more precise and rapid targeting. As AI models require vast amounts of data to refine their predictions, companies are incentivized to collect even more user information, creating a self-perpetuating cycle of surveillance. This amplified data collection and processing capacity means that vulnerable individuals, particularly problem gamblers, can be more easily identified and exploited for profit. The opaque nature of AI, often referred to as a "black box," further complicates efforts to understand or control the data points that drive these targeting decisions, pushing for continuous and broader data acquisition.

This development underscores how first-party data, collected directly by companies, can be weaponized for predatory advertising, challenging policy solutions focused solely on third-party data restrictions. The ability of platforms like DraftKings to leverage their internal user data with AI means that even in a future with limited third-party cookies, targeted behavioral advertising can persist and potentially intensify. This raises critical questions about the efficacy of current privacy regulations and highlights the need for broader bans on behavioral advertising to curb the underlying incentives for data exploitation and the harms it enables.