Online sports betting company DraftKings is actively deploying machine learning models to identify and retarget losing gamblers with personalized promotions designed to drive further wagers.
The Mechanics of Predictive Loss Targeting
Modern ad tech relies on continuous data ingestion. DraftKings trains its machine learning model entirely on first-party data—metrics captured directly from its users. By evaluating historical betting patterns, the system isolates users who consistently sustain financial losses. Rather than implementing risk-mitigation protocols for these vulnerable segments, the algorithm flags them as targets for advertising.
The system subsequently dispatches targeted promotions designed to lure them back to the site to place more bets. These automated hooks target problem gamblers—individuals who repeatedly gamble despite harm to themselves, their finances, and their relationships. DraftKings maintains a direct commercial incentive to sustain this cycle, as these are the users actually making DraftKings money.
Amplifying Surveillance Harms Through Machine Learning
Online behavioral advertising has long incentivized aggressive data collection, but the integration of artificial intelligence fundamentally shifts the threat matrix. AI operates as a black box. Because human engineers cannot easily predict which data points a model will prioritize, the architectural imperative defaults to continuous data collection.
AI processes enormous data sets much faster than previous methods. This acceleration supercharges the harms of behavioral advertising. The infrastructure built to serve targeted gambling promotions does not exist in a vacuum. It feeds the broader surveillance industry.
Data harvested for targeted ad placement is being sold to insurance companies, banks, and state and federal government law enforcement agencies. Immigration and Customs Enforcement (ICE) and Customs and Border Protection (CBP) have turned to commercial big data and ad tech providers to support investigative operations, as evidenced by a Request for Information published by ICE earlier this year.
Why First-Party Data Exemptions Fail Policy Solutions
A critical technical detail in the DraftKings deployment is its reliance exclusively on first-party data. The company is not buying any additional data from third parties to fuel its machine learning model because its own platform generates the behavioral footprint for its users.
This reality exposes a flaw in current regulatory frameworks. Policy solutions that focus strictly on restricting third-party data sharing and selling do nothing to stop companies from weaponizing the data they collect directly from their own customers. If a platform owns the digital storefront, its internal machine learning stack can profile, target, and exploit user vulnerability without ever crossing corporate boundaries.
Under a comprehensive ban, companies stripped of the ability to serve personalized ads would lose the economic incentive to collect the behavioral data powering them.
Reclaiming User Agency
The deployment of predictive targeting by DraftKings illustrates how ad tech evolves and how companies find new ways to use our data against us. As automated profiling becomes more sophisticated, technical countermeasures remain critical for individual protection.
Users seeking to limit their exposure to algorithmic tracking can use digital self-defense tooling. Resources such as the EFF’s Surveillance Self-Defense project provide practical instructions for protecting yourself on mobile apps and on websites to mitigate the telemetry harvesting that powers modern behavioral advertising networks.
