← Back to digest
HN Threads

DraftKings reportedly targets customers likely to lose

What's this about?

EFF says DraftKings uses betting records to identify customers likely to place losing bets and sends them personalised promotions. The concern is that the same model may preferentially re-engage people showing signs of gambling harm.

What exactly is first-party behavioural targeting?

The company uses activity collected directly inside its own service rather than buying third-party data. Limits on data brokers alone would therefore not stop it.

What's the community saying?

HN 529 points, 377 comments at selection time.
  1. Many compared betting apps with addictive drugs available continuously on a phone.
  2. Commenters argued this is a concrete AI harm suitable for regulation now.
  3. Others noted the optimisation is not technically novel; the ethical issue is the incentive and target, not the label “AI.”
← Thariq Shihipar describes mutable agent harnessesAI may need $6 trillion in annual revenue by 2031 →