The Impact of AI on CSAM Offending and Deterrence Efforts

Artificial intelligence is rapidly becoming a powerful tool on both sides of the online safety equation. It is being exploited to create vast quantities of new and synthetic child abuse material (CSAM), leveraging everything from previous victims' images to stolen social media photos to create deepfakes. However, the same technology offers opportunities to law enforcement: AI is enabling automated analysis of immense datasets to identify victims and perpetrators, with the prospect of reducing the psychological harm to human investigators.

Pivoting from enforcement to prevention, the presentation will also highlight the critical need for "Safety by Design" principles to be embedded in online platforms. A key example of deterrence is the reThink chatbot, a partnership between the Internet Watch Foundation, the Lucy Faithfull Foundation, and Aylo (which operates Pornhub), and subsequently evaluated by the University of Tasmania. The chatbot demonstrated a significant deterrent effect by confronting users searching for CSAM. The University of Tasmania is co-leading a new initiative, the CSAM Deterrence Centre, in collaboration with Stop It Now (Australia) to work with tech partners on implementing deterrence messaging and other measures.

SPEAKER: Dr Scanlan is a Senior Lecturer at the University of Tasmania. He specialises in cybersecurity and privacy, having undertaken a PhD in the area and has been teaching and consulting with the industry for 20 years. His primary research interests relate to the technical and social harms of the internet, with a particular focus on child exploitation prevention, cybersecurity and machine learning. His research involves collaborating closely in multiple-disciplinary teams from psychology and law. He is an active researcher on projects relating to child exploitation prevention, data privacy, maritime cybersecurity and machine learning.