Prince Edward Island RCMP investigations into online child sexual abuse now include a significant portion of material generated by artificial intelligence, according to RCMP officials. Senior communications adviser Scott Ferris highlighted a noticeable rise in AI-generated child sexual exploitation material within RCMP jurisdictions on the Island. It is estimated that approximately one-third of RCMP online child abuse cases involve AI-generated content, with exclusions for Charlottetown, Summerside, and Kensington, which have independent police forces.
Const. Tim Keizer of P.E.I. RCMP noted that AI is increasingly used to create or alter images of young individuals, making it challenging to differentiate between real and AI-generated imagery due to the high level of realism. Under Canadian law, AI-generated child sexual abuse material is treated equally to actual depictions, as outlined in Section 163.1 of the Criminal Code, which prohibits any material depicting minors for sexual purposes.
Cpl. Gavin Moore mentioned in a July email that roughly two RCMP cases per month involving child sexual abuse imagery lead to charges. However, this figure likely represents less than half of the total cases received due to factors like insufficient evidence. AI models are estimated to be involved in around one-third of cases, either creating fake images and videos of non-existent children or producing “deepfakes” – digitally altered versions of real individuals. Deepfakes are commonly used in sextortion cases, where criminals use threats to extort money or explicit content from victims.
Last year, P.E.I. RCMP reported 26 instances of sextortion, with Summerside police identifying one case involving AI-generated child abuse material this year. Moreover, Charlottetown police also found one instance of AI-generated child abuse material in the current year. Keizer emphasized the need for educating young people and parents about the risks of sharing images online to promote social responsibility with technology.
According to Jacques Marcoux from the Canadian Centre for Child Protection, even AI-generated material not depicting real children can have significant negative impacts, as many images are based on the likeness of actual children in communities. Marcoux stressed the importance of monitoring AI-generated exploitation to gauge the scale of the issue and address the growing demand for such content. Despite the legal consistency in treating AI-generated material, tracking and addressing this problem remain crucial for understanding and combating its proliferation.