Machine learning model identifies payments for live streamed child abuse

New research from the Australian Institute of Criminology used a machine learning approach to produce a proof of concept model for identifying financial indicators associated with live streaming of child sexual abuse (CSA):

  • Comparing CSA live streaming transactions with a comparison sample of non-CSA live streaming transactions, this model was successful at identifying those who live streamed CSA, while making few errors in identifying those who did not.
  • Seven financial risk indicators were identified. Six risk indicators centred on the value of transactions, and one on the age of the individual making the transactions.
  • These findings reveal an important opportunity to use financial transactions as an avenue for detecting and disrupting CSA live streaming.

View the report