The **New York Times** recently reported on **DraftKings'** controversial strategy: employing machine learning models trained on customer betting records to pinpoint 'losing gamblers.' Once identified, these individuals are subjected to targeted advertising campaigns designed to entice them back to the platform, specifically to place more bets that **DraftKings** anticipates will be unprofitable for the user. ### AI Supercharges Predatory Advertising While online behavioral advertising is not new, the integration of AI magnifies its potential harms. AI models can process vast datasets at unprecedented speeds, allowing companies like **DraftKings** to refine their targeting with alarming precision. This creates a strong business incentive to continuously collect more data, as the 'black box' nature of AI often means developers are unsure which data points will prove most valuable for model refinement. For **DraftKings**, the motivation is clear: losing gamblers are their most profitable users. However, this strategy disproportionately affects 'problem gamblers' – individuals who continue to gamble despite severe personal, financial, and relational consequences. By re-engaging these vulnerable individuals through highly personalized promotions, **DraftKings** is seen as exploiting addiction for profit rather than promoting responsible gambling. ### The Broader Surveillance Implications Beyond individual harm, online behavioral advertising fuels a larger surveillance industry. Data collected for ad targeting is frequently sold to third parties, including insurance companies, banks, and even government law enforcement agencies like **CBP** (Customs and Border Protection) and **ICE** (Immigration and Customs Enforcement). **ICE**, for instance, has actively sought information from commercial Big Data and Ad Tech providers to support its investigative activities, as evidenced by a recent Request for Information. ### First-Party Data: A Loophole for Predatory Practices? **DraftKings** appears to be primarily utilizing 'first-party data' – information collected directly from its users – to power its AI models. This highlights a crucial point for policymakers: solutions focused solely on restricting third-party data sharing and sales may not be sufficient to curb predatory advertising practices. The **Electronic Frontier Foundation (EFF)** has long advocated for a complete ban on all online behavioral advertising, arguing that eliminating personalized ads would significantly reduce the incentive for companies to collect the vast amounts of behavioral data currently fueling this ecosystem. ### Taking Back Control The case of **DraftKings** serves as a stark reminder of how ad tech continuously evolves, finding new ways to leverage personal data against individuals. While the **EFF** continues to push for systemic changes, users can take steps to protect their data. Resources like the **EFF's Surveillance Self Defense project** offer practical tips for securing data on mobile apps and websites, including tools like **Privacy Badger**. Ultimately, the ethical implications of AI-driven behavioral targeting, especially in sensitive areas like gambling, underscore the urgent need for comprehensive privacy regulations that prioritize user protection over corporate profit.