DraftKings is using AI to behaviorally target chronic gamblers

summary

DraftKings uses machine learning models trained on first-party user data to identify chronic, losing gamblers and lure them back onto the platform with targeted promotions. Because the company relies solely on first-party data, traditional policy proposals targeting third-party data sharing fail to prevent this exploitation. The Electronic Frontier Foundation highlights this practice as an example of how artificial intelligence supercharges the harms of online behavioral advertising. Consequently, the organization argues that policymakers must ban all behavioral advertising entirely to eliminate the underlying data collection incentives.

key points

1. DraftKings' machine learning models specifically target losing gamblers, capitalizing on problem behavior because these high-loss users generate the majority of platform revenue.
2. Because DraftKings uses strictly first-party data to train these models, regulations banning third-party data broker sales are completely insufficient to stop the practice.
3. The Electronic Frontier Foundation uses this case to argue that all online behavioral advertising must be banned to eliminate corporate incentives for mass data harvesting.

who should read it

Privacy advocates, policymakers, and software engineers interested in the intersection of machine learning, ad tech, and platform exploitation.

discussion hostile